{"source":"https://github.com/SaneSanders/noemium/tree/main/src/content/tools","count":331,"tools":[{"slug":"21st-dev","name":"21st.dev","tagline":"Component marketplace and registry for React and modern stacks.","url":"https://21st.dev","category":"design","pricing":"freemium","price_note":"free tier (browse + 2 component copies/day); Builder $6/mo yearly, Builder + AI $15/mo yearly, Team $7.50/seat/mo yearly","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"situational","verdict_text":"A registry for installing community React components with one command. The open-source collection is handy for discovery, but quality and maintenance vary, so treat any block as vet-before-ship.","limitations":["Component quality and maintenance vary widely.","Discovery can feel noisy as the catalog grows.","Some useful blocks are gated behind pro plans."],"strengths":["Open-source registry with one-command install for community React components.","Useful for discovering pre-built blocks and accelerating early prototypes."],"use_for":["Teams that want to browse and install community components quickly.","Prototyping when you can accept community-driven quality."],"skip_when":["You need production-ready components without manual vetting.","Discovery noise outweighs the time saved.","The blocks you want sit behind the paid Builder/AI tiers."],"receipts":["https://21st.dev","https://github.com/21st-dev"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"404s","name":"404s","tagline":"Gallery of creative 404 error page designs.","url":"https://404s.design","category":"design","pricing":"free","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Browse this when you need a 404 that does not look like a dead end. It is a mood board, not a component library: copy the idea, then rebuild it yourself.","limitations":["No code export.","Niche scope; only useful once per project.","Curation quality varies."],"strengths":["Free gallery of 404 designs for projects that need a better error page.","Useful starting point when you want inspiration before building it yourself."],"use_for":["Finding a visual direction for a 404 before writing any code.","Quick inspiration when the default error page feels like an afterthought."],"skip_when":["You need a drop-in component or exported code for the 404 page.","The project needs design assets beyond a single error page."],"receipts":["https://404s.design","https://404s.design/about"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"aceternity-ui","name":"Aceternity UI","tagline":"Animated React and Next.js components with strong visual effects.","url":"https://ui.aceternity.com","category":"design","pricing":"freemium","price_note":"free components; All-Access Pass (annual $169 or lifetime $199) for premium blocks and templates","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"ship","verdict_text":"Go-to when you need a landing page that feels expensive. The animations are opinionated but high quality, and the copy-paste workflow fits modern React stacks well.","limitations":["Effects can be heavy on low-end devices.","Pro tier gates the most impressive components.","Accessibility and performance need manual review."],"strengths":["Go-to when a landing page needs to feel expensive without a custom motion team.","Opinionated animations that still copy-paste into modern React and Next.js stacks."],"use_for":["Marketing pages and product sites that need high-quality motion out of a component kit.","React teams that already ship Tailwind and want copy-paste effects, not a design-system rewrite."],"skip_when":["The audience is on low-end devices and the effects would tank performance.","You need the Pro-gated showpieces on a free budget.","You cannot spare a pass for accessibility and performance review."],"receipts":["https://ui.aceternity.com","https://github.com/aceternity"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"activepieces","name":"Activepieces","tagline":"Open-source automation — unlimited flows, agents, MCP; cloud credits or Community Edition on-prem.","url":"https://www.activepieces.com","category":"automation","pricing":"freemium","price_note":"Free $0 (daily credits, 1 user); Plus $16/mo billed yearly; Team $166/mo billed yearly; Ultimate custom; Community Edition self-host free (core flows only)","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"n8n-shaped, MIT-flavored: unlimited flows, AI steps, MCP, and an honest split between Community Edition (flows) and paid (agents/chat, SSO, Git sync). Plus at $16/yr-billed is the BYOK on-ramp. Credit math is 1 per flow run, more for hosted models. Not Zapier-scale connectors.","limitations":["Community Edition drops Agents/Chat, API access, SSO, and Git Sync.","Free cloud is daily credits with a hard cap; overage does not PAYG.","Hosted AI steps cost 2–20 credits unless you BYOK (then 1 credit).","Connector coverage is thinner than Zapier/Make; HTTP piece is the escape hatch."],"strengths":["Open-source automation with unlimited flows and an honest split between free Community Edition and paid features.","MIT license plus self-hosted Community Edition for teams that need data on-prem.","AI steps, MCP support, and a BYOK option to cut hosted model credit costs."],"use_for":["Teams that want n8n-style automation under an MIT license with self-hosting.","Workflows that mix SaaS plumbing with occasional AI steps.","Budget-conscious shops that can run Community Edition and bring their own model keys."],"skip_when":["You need Zapier-scale connector breadth out of the box.","Community Edition covers your flows but you also need Agents, Chat, SSO, or Git Sync.","You cannot manage self-hosting or BYOK model keys."],"receipts":["https://www.activepieces.com/pricing","https://github.com/activepieces/activepieces"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"adobe-firefly","name":"Adobe Firefly","tagline":"Generative image, vector and design tools inside Adobe's creative cloud.","url":"https://firefly.adobe.com","category":"image","pricing":"freemium","price_note":"free daily generations; paid plans from $9.99/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"The safest choice for commercial work thanks to Adobe's trained-on-licensed- assets claim. Image quality is solid, not bleeding-edge, and it shines when combined with Photoshop/Illustrator workflows.","limitations":["Output can look more \"stock\" than Midjourney or Flux.","Credit system is easy to burn through quickly.","Best features require a Creative Cloud subscription."],"strengths":["Safest commercial-image pick thanks to Adobe's trained-on-licensed-assets claim.","Sits inside Photoshop and Illustrator instead of a separate playground."],"use_for":["Client work that has to survive a legal review.","Designers already living in Creative Cloud who need generative fills, not a new tool."],"skip_when":["You want bleeding-edge looks closer to Midjourney or Flux.","Credits burn faster than the brief can iterate.","You refuse a Creative Cloud subscription for the features that matter."],"receipts":["https://firefly.adobe.com","https://helpx.adobe.com/creative-cloud/apps/generative-ai/generative-credits-faq.html"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"ai21","name":"AI21","tagline":"Enterprise AI platform with Jamba models, Maestro optimization, and agent execution strategies.","url":"https://ai21.com","category":"models-api","pricing":"freemium","price_note":"New accounts get a $10 credit for three months; usage is then charged per token monthly.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Billing information is required once the trial credit expires or is exhausted."],"receipts":["https://ai21.com","https://docs.ai21.com/docs/usage-cost"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"aider","name":"Aider","tagline":"Open-source pair programming in the terminal — git-native, works with any model.","url":"https://aider.chat","category":"coding","pricing":"free","price_note":"free, open source; bring your own API key","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"ship","verdict_text":"The terminal purist's agent. No IDE lock-in, automatic commits, and it works with almost any model — local ones and the cheap Chinese flagships included. Wire up a good model and this beats most paid IDEs for a fraction of the cost.","limitations":["Terminal-only UI; no inline diff experience.","Output quality depends entirely on the model you wire up.","Repo-map can miss context in very large monorepos."],"strengths":["Git-native terminal agent: automatic commits, no IDE lock-in.","Works with almost any model — local weights and cheap Chinese flagships included.","Open source, so the bill is the model, not a seat."],"use_for":["Pair-programming in the terminal on a repo you already commit to.","BYOK setups where you pick the model for the job."],"skip_when":["You want inline diffs inside an IDE.","The model behind it is weak — Aider will not save a bad route.","A very large monorepo where the repo-map misses context."],"receipts":["https://github.com/Aider-AI/aider"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"amazon-bedrock","name":"Amazon Bedrock","tagline":"AWS-managed platform giving API access to hundreds of foundation models plus agents and guardrails.","url":"https://aws.amazon.com/bedrock/","category":"models-api","pricing":"paid","price_note":"Pay-per-use per model/token; rate card varies by provider and modality. New AWS customers receive up to $200 in credits to try AWS AI.","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Bedrock is the pragmatic choice for AWS shops that need a single API gateway to many frontier and open-weight models. It adds enterprise features like Knowledge Bases, Guardrails, and AgentCore that are hard to bolt onto a raw API. The convenience comes with AWS complexity; per-model pricing is not always the cheapest, and the rate card is a maze.","limitations":["Pricing spans dozens of providers and modalities; forecasting cost across a multi-model stack is non-trivial.","AWS account setup, IAM, and compliance overhead make it overkill for small prototypes.","Some newer models require region-specific availability and approval gates."],"strengths":["Single AWS API gateway to hundreds of foundation models plus agents and guardrails.","Knowledge Bases, Guardrails, and AgentCore are enterprise features baked in.","Fits AWS shops that already own IAM, compliance, and billing workflows."],"use_for":["AWS-native teams that need one contract and one API for many models.","Workloads that need guardrails, knowledge bases, or agent orchestration alongside inference."],"skip_when":["You are a small prototype that does not need AWS account, IAM, and compliance overhead.","Per-model cost forecasting across a multi-model stack is non-negotiable.","Newer models you need are locked behind region-specific availability or approval gates."],"receipts":["https://aws.amazon.com/bedrock/","https://aws.amazon.com/bedrock/pricing/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"amazon-polly","name":"Amazon Polly","tagline":"Fully managed AWS service that converts text into lifelike speech in dozens of languages.","url":"https://aws.amazon.com/polly","category":"audio","pricing":"freemium","price_note":"Free tier lasts 12 months and varies by engine (e.g., 1M neural characters/month); Standard voices are $4 per 1M characters, Neural $16, Generative $30, Long-Form $100.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Free tier is limited to 12 months and varies by voice engine; Long-Form and Generative voices are significantly more expensive."],"receipts":["https://aws.amazon.com/polly","https://aws.amazon.com/polly/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"amazon-transcribe","name":"Amazon Transcribe","tagline":"Fully managed AWS speech foundation model that converts real-time or recorded speech into text.","url":"https://aws.amazon.com/transcribe","category":"audio","pricing":"freemium","price_note":"60 minutes/month free for 12 months; standard streaming is $0.01/min and batch $0.006/min, with add-ons for PII redaction and custom language models.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Free tier covers only 60 minutes per month for the first 12 months; add-ons like PII redaction cost extra."],"receipts":["https://aws.amazon.com/transcribe","https://aws.amazon.com/transcribe/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"amie","name":"Amie","tagline":"AI calendar and meeting-notes app that summarizes meetings and automates follow-ups without a bot.","url":"https://amie.so","category":"productivity","pricing":"paid","price_note":"Pro is €20/user/mo billed annually; Business is €40/user/mo billed annually with CRM integrations and custom branding.","free_tier":false,"open_source":false,"api":false,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Pro and Business plans are billed annually; CRM integrations and custom branding on shared notes are limited to Business."],"receipts":["https://amie.so","https://amie.so/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"amp","name":"Amp","tagline":"Sourcegraph's agentic coding product (not Cody) — subscription orbs plus linked ChatGPT/Grok usage.","url":"https://ampcode.com","category":"coding","pricing":"paid","price_note":"Megawatt $20/mo (750 orb-hours, $20 included agent usage); Gigawatt $200/mo; unconstrained usage billing","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Amp is the product Sourcegraph actually pushes; Cody is the skip. $20 Megawatt includes agent usage and a pile of \"orbs,\" and you can attach a ChatGPT or SuperGrok sub for extra tokens at no Amp markup. Closed, hosted, and still a second IDE habit. Cody remains a separate, enterprise-only card.","limitations":["No public free tier on the Amp pricing page.","Orb hours and included agent dollars are two different meters.","Max two subscriptions per human; orbs are not a resale compute pool.","Team workspace billing is manual — no auto-seat provisioning yet."],"strengths":["Sourcegraph's current agentic coding push, with Cody kept as a separate enterprise card.","$20 Megawatt bundles agent usage and orb-hours for a low entry price.","Attach an existing ChatGPT or SuperGrok subscription at no Amp markup."],"use_for":["Developers who want agentic coding inside a second IDE and already tolerate Sourcegraph.","Teams that already pay for ChatGPT or SuperGrok and want to reuse that spend.","Small teams that can live with manual workspace billing."],"skip_when":["You need a free tier to evaluate before buying.","You expect orbs to act as a shared resale compute pool across the org.","You need auto-seat provisioning for a growing team."],"receipts":["https://ampcode.com/pricing","https://sourcegraph.com/docs/cody"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"anthropic-api","name":"Anthropic API","tagline":"Claude models direct — the practitioner's pick for coding and long-form text.","url":"https://platform.claude.com","category":"models-api","pricing":"paid","price_note":"pay-as-you-go, from $1/1M input (Haiku 4.5); small free credits for new users","free_tier":false,"open_source":false,"api":true,"self_host":false,"models_used":["claude-opus-5","claude-sonnet-5","claude-haiku-4-5"],"model_routing":"locked","verdict":"ship","verdict_text":"Hard to beat for code and long-form text — prompt caching and batch pricing make heavy workloads cheaper than they look. Quality is still top-tier. But the author of this catalog is moving more workloads to Kimi, Qwen and GLM; the price gap is getting hard to justify.","limitations":["Capacity crunches at peak demand have caused 429s.","Fewer auxiliary tools than OpenAI's ecosystem.","The Opus tier is expensive."],"strengths":["Hard to beat for code and long-form text at the quality tier.","Prompt caching and batch pricing make heavy workloads cheaper than the sticker."],"use_for":["Production Claude calls when you need the lab API, not the chat app.","Long documents and coding jobs where caching actually fires."],"skip_when":["Peak-hour 429s would stall the product.","You need OpenAI's wider auxiliary tooling.","Opus-tier spend is the only way the job works and the budget cannot take it."],"receipts":["https://docs.anthropic.com"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"anyword","name":"Anyword","tagline":"Predictive AI copywriting platform that scores marketing copy before it ships.","url":"https://www.anyword.com","category":"writing","pricing":"paid","price_note":"7-day free trial (2,500 words); Starter $49/mo ($39/mo annually); Data-Driven $99/mo ($79/mo annually); Business and Enterprise custom","free_tier":false,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Best for performance-obsessed marketing teams who want a predicted score on every ad, email, and landing line before spending media budget. The catch: predictive scoring is only as good as the data you feed it, and the useful tiers (Business/Enterprise) jump to custom pricing.","limitations":["Predictive scoring quality depends on connected channel data and historical performance rows.","API, SSO, private LLM, and custom integrations are Enterprise-only.","No permanent free tier; entry trial is 7 days / 2,500 words.","Long-form drafts still need human editing and brand-voice tuning."],"strengths":["Predictive scoring on ad, email, and landing copy before media budget is spent.","Purpose-built for performance-obsessed marketing teams."],"use_for":["Marketing teams that want a predicted score on every piece of copy.","Paid-media workflows where copy performance data feeds the model."],"skip_when":["You do not have enough historical performance data for scoring to be useful.","You need API, SSO, private LLM, or custom integrations without an Enterprise deal.","You want a permanent free tier beyond the 7-day / 2,500-word trial."],"receipts":["https://www.anyword.com","https://www.anyword.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"argus","name":"Argus","tagline":"Agentic visual UI testing agent that explores your app like a real user and reports what breaks.","url":"https://github.com/argus-testing/argus","category":"automation","pricing":"freemium","price_note":"source-available; free for individuals and small teams; enterprise license required for 100+ members","free_tier":true,"open_source":false,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"Describe a test in plain English and Argus drives a Playwright browser through it, capturing timelines, screenshots, and pass/fail reasoning. The local-first build keeps data on your machine, while a hosted option removes setup. Good for teams whose coding agents are producing UI faster than manual QA can keep up.","limitations":["Requires a Gemini API key for real runs in the local build.","Source-available license, not OSI open source; enterprise use is paid.","Multi-agent pipeline can be slower than a scripted test for simple, repetitive flows."],"strengths":["Turns a plain-English test description into a Playwright browser run with screenshots and reasoning.","Local-first build keeps runs, screenshots, and reports on your machine.","Hosted option removes setup for teams that do not want to self-host."],"use_for":["Teams whose coding agents ship UI faster than manual QA can review.","Smoke-testing a web app without writing and maintaining selectors."],"skip_when":["You need an OSI open-source license today; the source-available license is not OSI.","Simple, repetitive flows where a scripted test would be faster than the agent pipeline.","You do not have a Gemini, OpenAI, or Kimi API key for real local runs."],"receipts":["https://github.com/argus-testing/argus","https://github.com/argus-testing/argus/blob/main/LICENSE"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"arize-phoenix","name":"Arize Phoenix","tagline":"Open-source platform for AI application tracing, evaluation, prompt engineering, and experiments.","url":"https://arize.com/phoenix","category":"dev-infra","pricing":"freemium","price_note":"Open-source self-host is free; Phoenix Cloud includes 2 free instances; Arize AX Pro is $50 per month; Enterprise is custom.","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Phoenix Cloud free tier is limited to two instances; Arize AX Free caps traces at 25,000 spans and retention at 15 days."],"receipts":["https://arize.com/phoenix","https://docs.arize.com/phoenix/","https://arize.com/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"asana","name":"Asana","tagline":"AI-powered work management platform that helps teams plan, track, and get things done across projects and goals.","url":"https://asana.com","category":"productivity","pricing":"freemium","price_note":"Personal plan is free for up to 2 users; Starter $10.99/user/mo annually and Advanced $24.99/user/mo annually.","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Personal free plan caps projects and teams at 2 users, and AI Studio credits are limited on lower tiers."],"receipts":["https://asana.com","https://asana.com/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"assemblyai","name":"AssemblyAI","tagline":"Speech-to-text and voice-agent APIs — Universal-3.5 Pro, streaming billed on socket time.","url":"https://www.assemblyai.com","category":"audio","pricing":"freemium","price_note":"$50 free credits (no card); Universal-3.5 Pro $0.21/hr async / $0.45/hr streaming; Voice Agent API $4.50/hr all-in; add-ons stack","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"A real STT rate card, not a homepage teaser. Async Pro is $0.21/hr; streaming Pro is $0.45/hr on WebSocket open time (idle counts). Voice Agent is $4.50/hr bundled. EU region is the same price. $50 credits get you a prototype. Not Deepgram, not Whisper on your GPU.","limitations":["Streaming bills session duration, not speech duration — leaving a socket open is a bill.","Multichannel audio is billed per channel (stereo = 2×).","Add-ons stack on the base rate; diarization/PII/medical are extra meters.","Free-tier streaming concurrency is 5 new streams/min vs 100 on pay-as-you-go."],"strengths":["Transparent per-hour rate card ($0.21 async / $0.45 streaming / $4.50 voice agent).","$50 free credits let you prototype without a card.","Voice Agent API bundles everything at one hourly rate."],"use_for":["Prototyping STT or voice agents on the $50 credit pool.","Async transcription jobs where per-hour cost is the main metric.","Real-time streaming when you can tolerate session-duration billing."],"skip_when":["You need streaming billed by speech time, not idle WebSocket time.","Multichannel audio would multiply the hourly rate beyond budget.","Free-tier streaming concurrency (5 streams/min) would be the production path."],"receipts":["https://www.assemblyai.com/pricing","https://www.assemblyai.com/docs"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"assistant-ui","name":"assistant-ui Elements","tagline":"Open-source React components for building AI chat interfaces.","url":"https://assistant-ui.com","category":"design","pricing":"freemium","price_note":"open-source MIT core; assistant-cloud free up to 200 MAU, Pro $50/mo","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"Saves a lot of time when you need a chat UI that looks professional without building it from scratch. The React integration is clean, though the cloud extras are not necessary for most projects.","limitations":["Cloud features can blur the open-core boundary.","Custom thread UI still requires effort.","Smaller ecosystem than general component libraries."],"strengths":["Open-source React chat components that look professional without building from scratch.","Clean React integration for dropping an AI chat interface into an existing frontend.","MIT core can be self-hosted if you want the conversation UI on your own infrastructure."],"use_for":["Teams that need a polished chat UI inside a React app without writing it from zero.","Projects where the open-source core is enough and cloud extras are unnecessary."],"skip_when":["You need a packaged chat widget with zero frontend customization.","Cloud-only features are required, since they blur the open-core boundary.","You rely on a large ecosystem of general-purpose component libraries."],"receipts":["https://assistant-ui.com","https://github.com/assistant-ui/assistant-ui"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"atlassian-mcp","name":"Atlassian MCP","tagline":"Jira/Confluence MCP so the coding agent can read tickets instead of pasting them into chat.","url":"https://www.atlassian.com/platform/remote-mcp-server","category":"mcp","pricing":"freemium","price_note":"Requires an Atlassian cloud site; MCP itself is not a separate SKU we found","free_tier":false,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: remote MCP + company Jira is a data-boundary decision, not a toy.","On-prem Jira is a different story — this card is cloud-shaped."],"receipts":["https://www.atlassian.com/platform/remote-mcp-server"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"augment-code","name":"Augment Code","tagline":"Enterprise AI coding platform with a semantic Context Engine for very large codebases and multi-agent orchestration.","url":"https://www.augmentcode.com","category":"coding","pricing":"paid","price_note":"Business $100/mo flat (up to 50 seats, $100 usage included); Enterprise custom. Overage: provider LLM pricing + 40% service fee + Cosmos compute $0.19/hr. Trial: 30,000 credits with a card on file; no perpetual free tier.","free_tier":false,"open_source":false,"api":false,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Enterprise platform for huge codebases with a Context Engine that indexes 400K+ files across repos and Cosmos/Intent multi-agent orchestration. Supports Claude, GPT-5.x, Gemini and Kimi with Prism routing. Fits teams with monorepos and hard security bars (SOC 2 Type II, ISO 42001, CMEK). Main catch is the $100/mo minimum after individual plans were retired, plus usage-based overage that makes the bill unpredictable with active agents.","limitations":["Minimum plan is $100/mo — individual plans (~$20-60) retired, out of reach for solo devs.","Usage-based billing (LLM + 40% fee + $0.19/hr compute) scales unpredictably.","Heavy onboarding — initial indexing of large codebases takes tens of minutes before the Context Engine pays off."],"strengths":["Context Engine indexes 400K+ files across repos for huge codebases.","Cosmos/Intent multi-agent orchestration with Prism routing across several models.","Enterprise security bars: SOC 2 Type II, ISO 42001, CMEK."],"use_for":["Large monorepos where context span is the bottleneck.","Teams with strict security and compliance requirements for AI coding tools."],"skip_when":["Solo developers or small teams hit by the $100/mo minimum.","Unpredictable usage-based overage would break the budget.","You need fast onboarding without tens of minutes of initial indexing."],"receipts":["https://www.augmentcode.com/pricing","https://docs.augmentcode.com/models/token-based-pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"autogen","name":"AutoGen","tagline":"Microsoft framework for building multi-agent LLM applications.","url":"https://github.com/microsoft/autogen","category":"agents","pricing":"free","price_note":"open source (MIT)","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"Multi-agent orchestration with human-in-the-loop, code execution, and group chat patterns. Still useful for research prototypes and complex reasoning chains, but now in maintenance mode and the debugging experience for agent conversations is rough.","limitations":["Now in maintenance mode; new users are directed to Microsoft Agent Framework.","Steep learning curve compared to CrewAI.","Documentation can lag the fast-moving codebase.","Production observability requires extra tooling."],"strengths":["Open-source Microsoft framework with human-in-the-loop and code execution.","Supports group chat patterns for complex reasoning chains.","Free under the MIT license."],"use_for":["Research prototypes that need multi-agent conversation patterns.","Complex reasoning chains before committing to a managed platform."],"skip_when":["You want active development rather than a maintenance-mode project.","Production observability and debugging are required out of the box.","The learning curve of a fast-moving codebase is a blocker."],"receipts":["https://github.com/microsoft/autogen","https://microsoft.github.io/autogen/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"redshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"azure-speech","name":"Azure AI Speech","tagline":"Microsoft speech-to-text, TTS, and translation APIs — boring, regional, and actually contractable.","url":"https://azure.microsoft.com/en-us/products/ai-foundry/tools/speech","category":"audio","pricing":"paid","price_note":"Azure meter; free trial credits exist on new accounts","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: we did not re-price the meter this week — open the Azure calculator.","Enterprise procurement is the feature; indie UX is not."],"receipts":["https://azure.microsoft.com/en-us/products/ai-foundry/tools/speech","https://learn.microsoft.com/azure/ai-services/speech-service/"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"baidu-comate","name":"Baidu Comate","tagline":"Baidu's AI coding assistant with completion, chat, and Zulu agent mode for Chinese developers.","url":"https://comate.baidu.com","category":"coding","pricing":"freemium","price_note":"Standard free for individuals; Pro starts at ¥1 for the first month then ¥60/mo (or ¥500/yr); Enterprise from ¥799/seat/yr; new signups receive ¥10 model credit","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"A sensible choice for Chinese-speaking teams already using Baidu's cloud stack and ERNIE models. The catch: the most useful agentic features sit behind Pro or Enterprise tiers, and non-Chinese users will hit language, payment, and support friction.","limitations":["Documentation and support are Chinese-first.","Agentic and advanced generation features require Pro or Enterprise.","Not open source; tied to the Baidu ERNIE model ecosystem.","Enterprise pricing starts at ¥799/seat/year; dedicated support only at 1000+ seats.","International availability and payment methods are unclear."],"strengths":["Tight fit for Chinese-speaking teams already on Baidu Cloud and ERNIE models.","Standard tier is free for individuals.","Includes Zulu agent mode for users on Pro or Enterprise tiers."],"use_for":["Developers working in Chinese who want an IDE assistant inside Baidu's stack.","Teams already paying for Baidu Cloud and ERNIE services."],"skip_when":["Your team needs English-first documentation and support.","You want agentic features without paying for Pro or Enterprise.","You need BYOK or an open model ecosystem."],"receipts":["https://comate.baidu.com/zh/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"bardeen","name":"Bardeen","tagline":"Browser-native automation with AI that scrapes and runs workflows in the cloud.","url":"https://www.bardeen.ai","category":"automation","pricing":"freemium","price_note":"free tier (100 credits/mo); paid from $10/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Excellent for browser-heavy tasks — scraping, LinkedIn outreach, meeting prep. The cloud scheduler removes the need to keep a machine awake, but advanced workflows can be brittle when sites change.","limitations":["Relies on browser selectors that break when websites redesign.","Advanced AI features require paid plans.","Less suited for backend/API-heavy automations."],"strengths":["Strong at browser-heavy tasks like scraping, LinkedIn outreach, and meeting prep.","Cloud scheduler removes the need to keep a machine awake for automations."],"use_for":["Browser-first repetitive work that can be expressed as a playbook.","Quick lead enrichment and CRM data entry without backend access."],"skip_when":["The workflow is backend/API-heavy rather than browser-driven.","Target websites redesign often and would break selectors frequently.","You need advanced AI features without paying for a higher tier."],"receipts":["https://www.bardeen.ai","https://www.bardeen.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"baseten","name":"Baseten","tagline":"Model serving and inference infrastructure for production ML.","url":"https://www.baseten.co","category":"dev-infra","pricing":"paid","price_note":"usage-based; new accounts get free credits; enterprise plans support self-host","free_tier":false,"open_source":false,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"Solid managed inference for custom models with scaling and observability. Strong for teams that outgrow Replicate but don't want to build their own serving stack. Pricing is opaque and requires sales for big deployments.","limitations":["No free tier for production use.","Pricing requires talking to sales for accurate estimates.","Smaller ecosystem than hyperscaler options."],"strengths":["Managed inference with scaling and observability for custom models.","Strong fit for teams that have outgrown Replicate.","Enterprise plans offer a self-host option."],"use_for":["Production ML models that need managed serving without building the stack.","Teams outgrowing Replicate but not ready to build their own serving.","Enterprises that need self-hosted inference for compliance."],"skip_when":["You need transparent self-serve pricing before talking to sales.","A smaller ecosystem than hyperscalers is a blocker.","You need a free tier for production workloads."],"receipts":["https://www.baseten.co","https://www.baseten.co/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"bolt-new","name":"Bolt.new","tagline":"StackBlitz's in-browser AI app builder — prompt to a hosted full-stack site, billed in tokens.","url":"https://bolt.new","category":"coding","pricing":"freemium","price_note":"Free $0 (300k tokens/day, 1M/mo, Bolt branding); Pro $25/mo (from 10M tokens/mo, rollover); Teams $30/member/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Fastest path from a sentence to a public URL, running in WebContainers. Token burn scales with project size because the filesystem is synced into context. Free is a demo with branding and a daily cap. Not a replacement for a real repo once the app outgrows the sandbox.","limitations":["Token cost tracks project size; big codebases eat the monthly bucket fast.","Free sites carry Bolt branding and a 10MB upload cap.","Teams tokens are per member, not a shared pool.","Generated apps still need a human pass for auth, data, and production ops."],"strengths":["Fastest path from a sentence to a public URL inside a browser.","Runs full-stack prototypes in WebContainers without local setup.","Free tier lets anyone demo the workflow before paying."],"use_for":["Quick prototypes and demo sites that can stay in the sandbox.","First-time builders who want a deployed URL without tooling setup."],"skip_when":["The project will outgrow a sandboxed filesystem.","Token burn on a large codebase would exceed the monthly bucket.","You need production-grade auth, data, or ops review."],"receipts":["https://bolt.new/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"braintrust","name":"Braintrust","tagline":"Active observability platform for instrumenting, understanding, and improving agents.","url":"https://www.braintrust.dev","category":"dev-infra","pricing":"freemium","price_note":"Starter is free with 1 GB processed data per month, 10,000 scored outputs per month, and 14-day retention; Pro is $249 per month; Enterprise is custom.","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Starter plan includes only 1 GB of processed data, 10,000 scored outputs, and 14-day data retention."],"receipts":["https://www.braintrust.dev","https://www.braintrust.dev/docs","https://www.braintrust.dev/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"brave-search-mcp","name":"Brave Search MCP","tagline":"Official-ish MCP wrapper around Brave Search API for agents that need web results without a browser.","url":"https://github.com/modelcontextprotocol/servers","category":"mcp","pricing":"freemium","price_note":"Server free; Brave Search API has a free tier then paid","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: API quota and current package name not re-verified beyond the servers repo.","Search quality ≠ Tavily-style LLM-ready snippets."],"receipts":["https://github.com/modelcontextprotocol/servers","https://brave.com/search/api/"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"browser-use","name":"Browser Use","tagline":"Open-source browser agent library plus managed cloud browsers — task-in, work-out, or raw CDP.","url":"https://browser-use.com","category":"agents","pricing":"freemium","price_note":"OSS library free; Cloud Free $0 (10 agent tasks/mo); PAYG credits; Dev $29/mo, Business $299/mo, Scaleup $999/mo; browser time $0.02/hour; default V4 model GPT-5.6 Luna $0.24/$1.44 per 1M tokens","free_tier":true,"open_source":true,"api":true,"self_host":true,"models_used":["gpt-5-6-luna"],"model_routing":"byok","verdict":"situational","verdict_text":"Two products, one brand: pip-install the agent locally, or buy managed browsers at $0.02/hour. Cloud Free is 10 agent tasks/month. Stealth benchmarks are vendor-published — treat as a receipt, not a guarantee. Adjacent to Browserbase, not a duplicate of it.","limitations":["Cloud Free is 10 agent tasks/month; PAYG starts after a credit top-up.","BYOK and custom proxies start at Dev ($29), not Free.","Residential proxy is $5/GB (Scaleup $4/GB) — stealth traffic is the real bill.","V2/V3 APIs still exist; new work should be V4 or you inherit old meters."],"strengths":["Open-source agent library with a local pip-install path and managed cloud option.","Two products, one brand covers both self-hosted and browser-as-a-service.","CDP/raw browser control is available."],"use_for":["Browser automation where you want open-source code you can audit.","Tasks that need managed browsers without building the infra yourself.","Stealth/scraping workflows where proxy costs are budgeted."],"skip_when":["Cloud Free's 10 tasks/month is too small for your volume.","You need BYOK or custom proxies but will not pay the Dev tier.","You treat vendor-published stealth benchmarks as guarantees."],"receipts":["https://browser-use.com/pricing.md","https://docs.browser-use.com/open-source/quickstart","https://github.com/browser-use/browser-use"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"browserbase","name":"Browserbase","tagline":"Serverless browser infrastructure for AI agents and scrapers.","url":"https://www.browserbase.com","category":"data","pricing":"freemium","price_note":"Free; Developer $20/mo; Startup $99/mo; Scale custom","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Reliable managed browsers for agents that need to click, fill forms, and capture pages. Pricing is straightforward, but you still write the agent logic yourself.","limitations":["Requires coding the agent/scraper on top.","Costs scale with browser session time.","Anti-bot protections on some sites still trigger blocks."],"strengths":["Reliable managed browsers for agents that click, fill forms and capture pages.","Straightforward tiered pricing with a free start."],"use_for":["AI agents and scrapers that need real browser interaction.","Teams that do not want to run their own browser fleet."],"skip_when":["You are not willing to write the agent/scraper logic yourself.","Session time would scale costs beyond what headless alternatives cost.","Anti-bot protections on your target sites are a hard blocker."],"receipts":["https://www.browserbase.com","https://www.browserbase.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"bytedance-seed","name":"ByteDance Seed","tagline":"ByteDance Seed models served through Volcano Engine and BytePlus, including Seed 2.1 Pro/Turbo.","url":"https://seed.bytedance.com/en/","category":"models-api","pricing":"paid","price_note":"Seed 2.1 Pro approx $0.83/$4.17 per 1M (CNY ¥6/¥30), Seed 2.1 Turbo ~$0.42/$2.08; Doubao Pro 32K $0.11/$0.28; no official USD rate card, figures are conversions","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Seed 2.1 Pro is a frontier-class model at aggressive prices, backed by ByteDance's infrastructure. For Western users the access path is murky: no official USD rate card, mostly Chinese docs, and signup through Volcano Engine or BytePlus.","limitations":["Proprietary closed weights — no self-host or open-weights fallback.","Access outside China requires Volcano Engine/BytePlus setup; documentation mostly Chinese.","USD pricing is approximate; official rate card is CNY-only.","EU/US data-sovereignty and trust concerns around ByteDance."],"strengths":["Seed 2.1 Pro prices undercut frontier Western APIs on a conversion basis.","Backed by ByteDance's own infrastructure and model family.","Doubao Pro 32K is even cheaper for long-context work."],"use_for":["Cost-sensitive workloads where CNY pricing and Volcano Engine setup are acceptable.","Chinese-market applications that already use ByteDance cloud services."],"skip_when":["You need an official USD rate card or English-first documentation.","EU/US data sovereignty or ByteDance trust is a procurement blocker.","You require open-weights or self-hosted fallback."],"receipts":["https://llmcostcalc.com/cloud/doubao","https://essamamdani.com/blog/seed-2-1-pro-turbo-bytedance-doubao-june-2026","https://tokencost.app/models/doubao-seed-2-1-pro"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"camel-ai","name":"Camel AI","tagline":"Open-source framework for multi-agent collaboration and role-playing.","url":"https://www.camel-ai.org","category":"agents","pricing":"free","price_note":"open source (Apache 2.0)","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"Strong research-oriented framework with role-playing agents and memory. Useful for academic and prototype work. Smaller ecosystem and less tooling than AutoGen or CrewAI.","limitations":["Smaller community and fewer integrations than CrewAI.","Production deployment path is less mature.","Role-playing setup can be verbose."],"strengths":["Research-oriented framework with role-playing agents and memory built in.","Modular primitives for agents, societies, RAG, and synthetic data."],"use_for":["Academic experiments and agent prototype work.","Multi-agent sketches where role-playing is central to the design."],"skip_when":["You need the integrations and community size of CrewAI or AutoGen.","You want a clear production deployment path out of the box.","Verbose role-playing setup would slow your iteration."],"receipts":["https://www.camel-ai.org","https://docs.camel-ai.org"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"canva-code","name":"Canva Code","tagline":"AI site/app builder inside Canva; prompt a front-end experience, then edit it like a design.","url":"https://www.canva.com","category":"design","pricing":"freemium","price_note":"free tier includes limited Premium AI uses; paid Pro/Business/Enterprise plans add more AI credits; AI Pass add-on available","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Good for marketers, educators, and small teams who want a pretty interactive landing page or mini app without leaving Canva. The catch: it is front-end only and not meant for complex backends or high-traffic sites; free credits are tight and paid AI allowances still meter usage.","limitations":["Free tier caps Premium AI uses (roughly 20/mo on the public pricing page).","Complex backends, authentication, and high-traffic scaling are out of scope.","Moving elements inside generated output still requires re-prompting.","Outputs live in Canva's ecosystem; exporting clean code is limited."],"strengths":["Builds interactive landing pages and mini apps inside Canva's familiar canvas.","Useful for marketers, educators, and small teams without a frontend engineer.","Free tier exists, though Premium AI uses are capped."],"use_for":["Quick marketing landing pages or simple interactive content.","Teams already paying for Canva Pro/Business and want AI-generated front-ends."],"skip_when":["You need a real backend, authentication, or high-traffic scaling.","You want clean exported code outside Canva's ecosystem.","Free credits are too tight for your expected volume."],"receipts":["https://www.canva.com","https://www.canva.com/pricing/","https://www.canva.com/newsroom/news/Canva-Code/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"capcut","name":"CapCut","tagline":"ByteDance's AI-powered photo and video editor for social content, with templates, effects, and generative tools.","url":"https://www.capcut.com","category":"video","pricing":"freemium","price_note":"Free online editor and mobile app; CapCut Pro subscription unlocks premium assets, effects, and higher-end exports (no public HTML price card found today — verify in-app at checkout).","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"CapCut is the fastest path to polished short-form video for TikTok, Reels, and YouTube, especially on mobile. The free tier covers most casual creators, while Pro adds premium templates, AI tools, and higher-quality exports. It is less suited to long-form or team workflows that need frame-level precision, version control, or enterprise data governance.","limitations":["ByteDance ownership and data-handling practices are a procurement concern for some businesses and institutions.","Advanced color grading, audio mixing, and collaboration lag behind desktop editors like DaVinci Resolve or Premiere.","AI generative features are convenient but quality varies; exported free projects may nudge users toward paid assets."],"strengths":["Fastest path to polished short-form video for TikTok, Reels, and YouTube.","Free tier covers most casual creators.","Mobile-first workflow with templates, effects, and generative tools."],"use_for":["Casual creators making short-form social content on phone or web.","Quick edits where templates and auto-effects beat manual timeline work."],"skip_when":["You need long-form editing, frame-level precision, or team version control.","ByteDance data handling is a procurement blocker for your business.","Advanced color grading, audio mixing, or desktop-editor control is required."],"receipts":["https://www.capcut.com","https://www.capcut.com/tools/ai-video-editor"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"cartesia","name":"Cartesia","tagline":"Ultra-low-latency voice AI with streaming TTS and voice cloning.","url":"https://cartesia.ai","category":"audio","pricing":"freemium","price_note":"free tier (20K credits/mo); paid plans from $5/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"ship","verdict_text":"Sonic models are built for speed: sub-100 ms streaming makes them excellent for voice agents and real-time apps. Voice cloning is strong, and the free tier is now usable for light projects; paid plans start at $5/mo.","limitations":["Free tier credit cap limits sustained production use.","Voice library is smaller than ElevenLabs.","Usage-based pricing requires careful cost forecasting."],"strengths":["Sonic models stream under 100 ms — built for voice agents, not studio takes.","Voice cloning is strong, and the free tier is now usable for light projects."],"use_for":["Real-time voice agents where latency is the product.","Light prototypes before you commit to usage-based spend."],"skip_when":["You need a large voice library closer to ElevenLabs.","Sustained production would blow the free credit cap.","You cannot forecast usage-based TTS cost."],"receipts":["https://cartesia.ai","https://cartesia.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"cerebras","name":"Cerebras","tagline":"Ultra-fast inference on wafer-scale AI accelerators with competitive API pricing.","url":"https://cerebras.ai","category":"models-api","pricing":"freemium","price_note":"free trial ($5 credits); self-serve Developer from $10 deposit; Enterprise custom","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"ship","verdict_text":"Latency is the selling point: responses feel instant on big open models. Pricing undercuts many cloud APIs. The model menu is narrower and you need to trust their hardware stack.","limitations":["Model catalog is smaller than OpenRouter or Together.","Requires migration to their API endpoints.","Long-term roadmap depends on hardware business."],"strengths":["Latency is the selling point: big open models feel instant.","API pricing undercuts many cloud inference hosts."],"use_for":["Autocomplete, voice, or chat UIs where wait time kills the loop.","Open models you already picked, if they are on the Cerebras menu."],"skip_when":["You need the wide catalog of OpenRouter or Together.","Migrating off your current endpoints is not worth the speed.","You will not bet the roadmap on their hardware business."],"receipts":["https://cerebras.ai","https://www.cerebras.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"chatgpt","name":"ChatGPT","tagline":"The general-purpose AI assistant that set the category baseline.","url":"https://chatgpt.com","category":"productivity","pricing":"freemium","price_note":"free tier; Go $8/mo, Plus $20/mo, Pro from $100/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["gpt-5-6-terra","gpt-5-4-mini","gpt-5-6-sol","gpt-5-6-luna"],"verdict":"ship","verdict_text":"Still the safest default assistant — sit it next to Claude and Gemini, not Grammarly. Reasoning, search, code execution and images in one place; the free tier still beats many paid rivals. Image generation is a separate catalog job (GPT Image 2); Codex is a separate coding card.","limitations":["Usage caps on Plus frustrate heavy users mid-task.","Memory and personalization features can go stale or wrong.","The app keeps pushing features you may not want.","This is the chat assistant, not the API and not Codex."],"strengths":["Safest default assistant: reasoning, search, code execution and images in one place.","The free tier still beats many paid rivals for everyday work."],"use_for":["Daily chat when you want one app, not a writing-only grammar tool.","Light research and reasoning where Plus or even free is enough."],"skip_when":["Plus caps cut you off mid-task.","You need the API or Codex — those are separate cards.","You want a quiet app that does not keep pushing extra features."],"receipts":["https://chatgpt.com/pricing","https://help.openai.com/en/articles/11369540","https://platform.openai.com/docs/models"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":true,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"chatoss","name":"ChatOSS","tagline":"Native desktop workspace for agentic coding with local Ollama models and optional cloud models.","url":"https://chatoss.ai/","category":"coding","pricing":"freemium","price_note":"Free for local Ollama; Starter $8/mo, Plus $20/mo, Pro $60/mo, Ultra $200/mo for cloud models","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"byok","verdict":"situational","verdict_text":"A Codex-style workspace built around open-weight models, with agents that write code, manage a Kanban board, and review PRs. The free tier is genuinely usable with local Ollama, and paid tiers add concurrent cloud models. Worth a look if you want an agentic IDE that does not default to OpenAI.","limitations":["Proprietary app; no public source repository yet.","Cloud plans require an account and ongoing payment; local-only mode needs a capable machine.","Newer than the established agentic IDEs, so ecosystem and stability are unproven."],"strengths":["Native desktop workspace built around open-weight models.","Free tier is genuinely usable with local Ollama.","Agents handle coding, Kanban, and PR review in one place."],"use_for":["Teams that want an agentic IDE defaulting to open weights, not OpenAI.","Coding with local Ollama and optional cloud model fallback."],"skip_when":["You need a mature plugin ecosystem and proven stability.","Your machine cannot run local models comfortably.","Full source transparency is required because the app is proprietary."],"receipts":["https://chatoss.ai/","https://chatoss.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"chroma","name":"Chroma","tagline":"Open-source search infrastructure for AI — local library plus usage-priced Cloud.","url":"https://www.trychroma.com","category":"data","pricing":"freemium","price_note":"OSS self-host free; Cloud Starter $0 + usage ($5 credits); Team $250/mo + usage ($100 credits, no rollover); write $2.50/GiB, storage $0.33/GiB-mo","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"The Python-native vector store people actually start with. Local Chroma is free. Cloud Starter is $0 plus $2.50/GiB write after a small credit pile; Team is a $250 seat on the same meters. Not Pinecone (that's managed-only) and not a retrieval API (that's Voyage). Untested at production QPS here.","limitations":["Starter is usage-priced; large ingest bills write GiB, not \"free forever\".","Team's $100 included credits do not roll over.","Hitting customer or Chroma usage limits pauses the service until raised.","Local OSS and Cloud are not feature-identical (serverless search vs laptop DB)."],"strengths":["Python-native vector store with a free local start.","Open-source self-host option keeps data on your own hardware.","Clear Cloud usage meters ($2.50/GiB write, $0.33/GiB-mo storage)."],"use_for":["Starting vector search locally in a Python project.","Teams that need a self-hosted vector store for RAG.","Workloads where local OSS and managed Cloud are both viable."],"skip_when":["You need a fully managed service like Pinecone with no local ops.","Large ingest would bill write GiB faster than a flat plan.","Feature parity between local OSS and Cloud is required."],"receipts":["https://www.trychroma.com/pricing","https://github.com/chroma-core/chroma"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"chrome-devtools-mcp","name":"Chrome DevTools MCP","tagline":"Official Chrome DevTools MCP — control and inspect a live Chrome from a coding agent.","url":"https://github.com/ChromeDevTools/chrome-devtools-mcp","category":"mcp","pricing":"free","price_note":"Apache-2.0; npx chrome-devtools-mcp@latest; Node 22+ and a current Chrome","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"ship","verdict_text":"The Chrome team's own MCP: traces, network, console, heap, a live debugging session. Complementary to Playwright MCP (this inspects Chrome, that drives any browser via snapshots). Needs a real Chrome and Node 22+.","limitations":["Requires Node 22+ and a current stable Chrome, not just any Chromium.","stdio server with the privileges of the user — a live debugging port is a live debugging port.","Tool catalog is large (heap, PWA, memory) and easy to over-enable."],"strengths":["Official Chrome DevTools team server — traces, network, console, heap.","Can attach to a running Chrome instead of spawning a throwaway profile.","Apache-2.0, npm chrome-devtools-mcp, published as io.github.ChromeDevTools."],"use_for":["Coding agents that need to debug or profile a live Chrome session.","Performance work (LCP, traces) that Playwright snapshots cannot see."],"skip_when":["You only need to click through a page — Playwright MCP is the thinner tool.","You cannot run current Chrome and Node 22+ on the machine.","You refuse to give an agent DevTools-level access to a browser profile."],"receipts":["https://github.com/ChromeDevTools/chrome-devtools-mcp","https://developer.chrome.com/docs/devtools/agents/get-started"],"affiliate":"none","evidence_tier":"field-tested","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"claude","name":"Claude","tagline":"Anthropic's assistant — the practitioner's favorite for prose and long documents.","url":"https://claude.ai","category":"productivity","pricing":"freemium","price_note":"free tier; Pro $20/mo ($17/mo billed yearly, $200 up front); Max from $100/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["claude-opus-5","claude-sonnet-5","claude-haiku-4-5"],"verdict":"ship","verdict_text":"The cleanest prose and the best long-document workflow of any chat assistant; Projects and Artifacts fit real work. Compare it with ChatGPT and Gemini on this shelf, not with Grammarly. Claude Code is a separate coding card and shares the same usage pool on paid plans.","limitations":["Usage caps on Pro hit fast with large context; Code/Cowork draw from the same pool.","No image generation.","Smaller plugin/integration ecosystem than ChatGPT."],"strengths":["Cleanest prose and the best long-document workflow of the chat assistants.","Projects and Artifacts fit real work, not just a demo canvas."],"use_for":["Long briefs, editing, and document-heavy thinking.","A practitioner chat seat compared with ChatGPT and Gemini, not Grammarly."],"skip_when":["Pro usage plus Claude Code would share one pool and burn it.","You need image generation in the same app.","You depend on ChatGPT's larger plugin surface."],"receipts":["https://www.anthropic.com/pricing","https://claude.ai"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"claude-code","name":"Claude Code","tagline":"Anthropic's terminal agent that edits, runs and tests code in your repo.","url":"https://docs.anthropic.com/en/docs/claude-code","category":"coding","pricing":"paid","price_note":"included in Claude Pro $20/mo ($17/mo yearly) and Max 5x $100 / Max 20x $200; shares the plan usage pool; extra usage at API rates","free_tier":false,"open_source":false,"api":true,"self_host":false,"models_used":["claude-opus-5","claude-sonnet-5","claude-haiku-4-5"],"model_routing":"locked","verdict":"ship","verdict_text":"Strong agentic coding, nobody argues with that — multi-file refactors, test runs and git workflows hold up. But the API bill on a big repo climbs fast, and it rewrites code with a confidence it hasn't always earned. The Chinese tools (Kimi, Qwen Code) are catching up at a fraction of the price, and that's where things are heading.","limitations":["API-billed sessions on large repos escalate in cost fast.","Confidently makes sweeping changes; big diffs need human review.","No GUI at all — a feature until you want a diff viewer.","Usage is shared with claude.ai chat on the same seat."],"strengths":["Multi-file refactors, tests, and git workflows that hold up in a real repo.","Best-in-class agentic coding on Anthropic models.","Lives in the repo — no forked IDE to adopt."],"use_for":["Terminal-native agents on a Claude seat you already pay for.","Large refactors you will still review before merge."],"skip_when":["Cost-sensitive work on a big repo — the API bill climbs fast.","You want a GUI diff viewer.","You want the Chinese-price stack (Kimi / Qwen) instead of Anthropic."],"receipts":["https://github.com/anthropics/claude-code","https://www.anthropic.com/pricing","https://docs.anthropic.com/en/docs/claude-code"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":true,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"clickup","name":"ClickUp","tagline":"All-in-one productivity platform with AI agents and 100+ tools to replace fragmented work software.","url":"https://clickup.com","category":"productivity","pricing":"freemium","price_note":"Free Forever plan includes unlimited tasks and members with 60 MB storage; Unlimited, Business, and Enterprise plans add views, automations, and governance.","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Free Forever plan storage is capped at 60 MB and guests must be given full access with no permission controls."],"receipts":["https://clickup.com","https://clickup.com/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"clinch","name":"Clinch","tagline":"Local-first macOS fork of Warp that keeps Claude Code and Codex sessions alive across restarts.","url":"https://clinch.sh/","category":"coding","pricing":"free","price_note":"AGPL-3.0 open source; no account, no telemetry","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"situational","verdict_text":"An unofficial Warp fork tuned for AI coding agents: project tabs, durable workspaces, and session capture so Claude Code and Codex conversations survive app restarts. Remote Control lets you answer a waiting agent from your phone over your own Tailscale network. macOS 14+ only and not notarized, so expect a Gatekeeper prompt.","limitations":["macOS 14+ only; no Linux or Windows build.","Public preview is ad-hoc signed, not notarized.","Session recovery is best-effort; provider transcripts and process state can still be lost."],"strengths":["Unofficial Warp fork built to keep Claude Code and Codex sessions alive across restarts.","Project tabs, durable workspaces, and session capture for agent workflows.","Remote Control answers a waiting agent from your phone over your own Tailscale network."],"use_for":["macOS developers running Claude Code or Codex who keep losing session state.","Teams that want agent sessions on a self-hosted, AGPL terminal without telemetry."],"skip_when":["You are on Linux or Windows; macOS 14+ is the only supported platform.","You cannot accept an ad-hoc signed, non-notarized Gatekeeper prompt.","Session recovery must be guaranteed rather than best-effort."],"receipts":["https://clinch.sh/","https://github.com/elliot-ylambda/clinch-terminal"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"cline","name":"Cline","tagline":"Open-source autonomous coding agent for VS Code and CLI — BYOK, you pay the model bill.","url":"https://cline.bot","category":"coding","pricing":"freemium","price_note":"extension/CLI free for individuals (BYOK or Cline inference at cost); optional ClinePass $9.99/mo for open-weight models; Enterprise custom (SSO, JetBrains, SLA)","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"The honest Cursor alternative if you already have API keys: Apache-licensed agent in the editor, MCP marketplace, no seat fee for solo use. The bill is the model, and agent loops on Sonnet/Opus burn tokens fast. Enterprise is a different product (JetBrains, SSO, provider allowlists). Not field-tested here.","limitations":["Inference is usage-based; a long agent session can exceed a $20 IDE sub.","You operate the keys, spend caps, and model routing yourself.","JetBrains, SSO, RBAC, and SLA are Enterprise-only on the public pricing page.","ClinePass is an optional paid subscription for curated open-weight models; the open-source editor/CLI remains free.","Client-side agent still needs you to review diffs; autonomy is not a guarantee."],"strengths":["Apache-licensed agent in the editor: no seat fee for solo use if you bring keys.","Honest Cursor alternative when the bill should be the model, not the IDE.","MCP marketplace without a second closed runtime."],"use_for":["VS Code / CLI agents on BYOK, with you watching the diff.","Solo use where ClinePass and Enterprise are optional, not the product."],"skip_when":["A long agent loop on Sonnet/Opus would exceed a $20 IDE sub.","You do not want to operate keys, spend caps, and routing yourself.","You need JetBrains, SSO, RBAC, or an SLA — that is Enterprise, not the OSS client."],"receipts":["https://cline.bot/pricing","https://cline.bot/cline-pass","https://github.com/cline/cline"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"clipdrop","name":"Clipdrop","tagline":"Jasper's suite of generative image tools — uncrop, relight, remove background, and more.","url":"https://clipdrop.co","category":"image","pricing":"freemium","price_note":"free tier with watermarks; Pro from $9/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Handy modular tools for quick edits: remove background, upscale, relight, inpaint. Individual features work well, but the suite feels fragmented and the best quality sits behind the paid tier.","limitations":["Free exports are watermarked at higher resolutions.","Multiple tools rather than one cohesive editor.","Model updates have been slower than Midjourney/Flux rivals."],"strengths":["Modular tools for quick edits like background removal, relight and inpaint.","Free tier exists to test before paying.","Paid Pro unlocks high-resolution, unwatermarked exports."],"use_for":["One-off image edits when you don't need a full editor.","Quick prototyping across multiple generative image tasks."],"skip_when":["You need one cohesive editor rather than fragmented tools.","Free watermarks at higher resolution break your deliverable.","You need the latest model quality competing with Midjourney/Flux."],"receipts":["https://clipdrop.co","https://clipdrop.co/remove-background","https://clipdrop.co/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"cloudflare-mcp","name":"Cloudflare MCP","tagline":"Cloudflare MCP servers for Workers/docs/builds — agent access to the account, not just docs search.","url":"https://developers.cloudflare.com/agents/model-context-protocol/","category":"mcp","pricing":"free","price_note":"MCP is free; Cloudflare usage is the bill","free_tier":true,"open_source":true,"api":true,"self_host":false,"limitations":["Radar: account-scoped MCP can mutate production Workers.","Docs URL is the receipt; we have not run every Cloudflare MCP server in this wave."],"receipts":["https://developers.cloudflare.com/agents/model-context-protocol/"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"cloudflare-vibesdk","name":"Cloudflare VibeSDK","tagline":"Open-source platform for self-hosting a full-stack vibe-coding environment on Cloudflare.","url":"https://github.com/cloudflare/vibesdk","category":"dev-infra","pricing":"free","price_note":"open source (MIT); production previews require Cloudflare Workers Paid plan, Workers for Platforms, model API keys, and a custom domain/certs","free_tier":true,"open_source":true,"api":false,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"For teams that want to own their vibe-coding stack and already run on Cloudflare. The catch: production features need a paid Cloudflare plan, multiple bindings (KV, D1, R2, AI Gateway), and you bring your own model keys; local setup is straightforward but production ops are not trivial.","limitations":["Production app previews require Cloudflare Workers Paid plan and Workers for Platforms.","You must configure AI Gateway / model provider keys yourself.","Bash is disabled for safety, but agent loops can still produce incorrect or insecure code.","Self-hosted means you operate DNS, certs, migrations, and quotas."],"strengths":["Open-source full-stack vibe-coding environment.","Self-hosted on your own Cloudflare account.","Local setup is straightforward."],"use_for":["Teams already running on Cloudflare who want a self-hosted vibe stack.","Developers who need to own their coding-assistant infrastructure."],"skip_when":["You do not want a paid Cloudflare plan for production previews.","You cannot manage KV, D1, R2, AI Gateway bindings and keys.","Bash in the agent loop is a hard requirement."],"receipts":["https://github.com/cloudflare/vibesdk","https://github.com/cloudflare/vibesdk/blob/main/docs/setup.md"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"cloudflare-workers-ai","name":"Cloudflare Workers AI","tagline":"Serverless GPU inference for 50+ open-source models on Cloudflare's global network, with free and paid plans.","url":"https://developers.cloudflare.com/workers-ai/","category":"models-api","pricing":"freemium","price_note":"Pay-for-what-you-use per-token/neuron rates; a free plan is available with limits.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Model catalog is a curated set of 50+ open-source models, not the full Hugging Face hub."],"receipts":["https://developers.cloudflare.com/workers-ai/","https://developers.cloudflare.com/workers-ai/platform/pricing/"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"codegeex","name":"CodeGeeX","tagline":"Free and open-source IDE plugin and coding model family from Zhipu.","url":"https://codegeex.cn","category":"coding","pricing":"free","price_note":"CodeGeeX plugin and model family are free and open source; GLM Coding Plan paid subscription is a separate product from $18/mo","free_tier":true,"open_source":true,"api":false,"self_host":true,"model_routing":"locked","verdict":"ship","verdict_text":"A genuinely free, open-source Copilot-style plugin with self-hosting options. It will not match Cursor or GitHub Copilot in ecosystem depth, but the price and privacy story are hard to beat.","limitations":["Ecosystem smaller than Cursor or GitHub Copilot.","Western documentation and support are thinner.","Model updates and roadmap tied to Zhipu/GLM.","Enterprise features live in the separate paid GLM Coding Plan."],"strengths":["Genuinely free, open-source Copilot-style plugin with a self-host path.","Price and privacy story are hard to beat even if the ecosystem is thinner."],"use_for":["Teams that want inline completions without a Copilot invoice.","Air-gapped or self-hosted coding assistance on GLM-class models."],"skip_when":["You need Cursor or Copilot ecosystem depth.","Western docs and support are a requirement.","Enterprise features that only exist on the paid GLM Coding Plan."],"receipts":["https://codegeex.cn","https://github.com/CodeGeeX","https://docs.z.ai/devpack/overview","https://docs.z.ai/guides/overview/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"codesandbox","name":"CodeSandbox","tagline":"Instant, isolated cloud sandboxes for code execution, development environments, and AI-agent runtimes.","url":"https://codesandbox.io","category":"dev-infra","pricing":"freemium","price_note":"Build plan is free with 5 members, 20 GB storage, 20 new SDK sandboxes per hour, and 10 concurrent VM sandboxes; Pro, Scale, and Enterprise are paid.","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Build plan caps members at 5, new SDK sandboxes at 20 per hour, and concurrent VM sandboxes at 10."],"receipts":["https://codesandbox.io","https://codesandbox.io/docs","https://codesandbox.io/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"cody","name":"Sourcegraph Cody","tagline":"AI coding assistant with deep code-search context from Sourcegraph.","url":"https://sourcegraph.com/cody","category":"coding","pricing":"paid","price_note":"enterprise-only (self-serve discontinued 2025-07); contact sales","free_tier":false,"open_source":false,"api":false,"self_host":false,"model_routing":"bundled","verdict":"skip","verdict_text":"Code-search context was a real edge once. Then Sourcegraph's attention moved to Amp, self-serve plans got killed, and Cody's release cadence shows where it sits in the priority list. Pick a tool its maker actually cares about.","limitations":["Self-serve Free/Pro discontinued; individuals are pointed to Amp instead.","Development focus shifted to Amp; Cody feels maintenance-mode.","Agentic features trail the category leaders."],"strengths":["Deep code-search context from Sourcegraph's codebase graph.","Enterprise-grade admin and security controls for large codebases."],"use_for":["Organizations already on Sourcegraph Enterprise that want bundled AI assistance.","Teams that need code-search context tied to an existing Sourcegraph deployment."],"skip_when":["You are an individual developer looking for self-serve plans.","You want a vendor whose current roadmap prioritizes the AI assistant.","Agentic features are a priority over legacy code search."],"receipts":["https://sourcegraph.com/docs/cody","https://sourcegraph.com/blog/changes-to-cody-free-pro-and-enterprise-starter-plans"],"affiliate":"none","evidence_tier":"source-verified","momentum":"redshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"cogni","name":"Cogni","tagline":"MCP memory server that connects facts in an entity graph so agents reason over what they remember.","url":"https://getcogni.io/","category":"agents","pricing":"freemium","price_note":"Free personal tier; Pro for higher limits and graph inspection; Team/Enterprise custom","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Exposes memory tools over MCP with deterministic graph traversal rather than pure vector similarity, which helps with multi-hop questions that share no vocabulary with the answer. The free tier is genuinely useful, and it works with any model. Hosted-only for now; self-hosted deployments are on the roadmap.","limitations":["Hosted service; memory text leaves your machine unless you wait for the self-hosted tier.","Effectiveness depends on the agent consistently calling recall before answering.","Pro and team pricing is not published transparently on the site."],"strengths":["Deterministic entity-graph traversal reaches connected facts a vector search misses.","Works with any model over MCP; no model lock-in.","Free personal tier is genuinely useful.","No vector database or separate model to run."],"use_for":["Agents that need multi-hop reasoning across disconnected facts.","Users who want persistent memory across sessions and clients."],"skip_when":["You cannot use a hosted service; self-hosted deployment is still on the roadmap.","Your agent will not consistently call recall before answering.","Published Pro/Team pricing is required before you buy."],"receipts":["https://getcogni.io/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"cohere","name":"Cohere","tagline":"Enterprise-focused models — Command for generation, Rerank for retrieval.","url":"https://cohere.com","category":"models-api","pricing":"freemium","price_note":"free trial keys (1,000 calls/mo); production API pay-as-you-go (Command from $1/1M input tokens)","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Rerank is the real product — it quietly improves almost any RAG pipeline, no matter which LLM you use. Command is fine, but it isn't why you sign up. Enterprise sales and confusing pricing don't help.","limitations":["Chat models trail frontier quality.","Enterprise-sales motion makes self-serve feel second-class.","Pricing pages for newer models are confusing."],"strengths":["Rerank quietly improves almost any RAG pipeline regardless of the LLM.","Free trial keys give 1,000 calls/mo to test before paying."],"use_for":["Boosting retrieval quality in RAG without switching your generator.","Enterprise search where reranking is the primary need."],"skip_when":["You need frontier chat-model quality from the same API.","Self-serve pricing and clear model pages are non-negotiable.","Enterprise sales motion would slow your procurement."],"receipts":["https://cohere.com","https://cohere.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"cohere-embed","name":"Cohere Embed","tagline":"Embedding service that turns text and images into vectors for semantic retrieval and agentic applications.","url":"https://cohere.com/embed","category":"data","pricing":"paid","free_tier":false,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Pricing is not listed on the Embed product page; usage is through the Cohere API."],"receipts":["https://cohere.com/embed"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"comet","name":"Comet","tagline":"Perplexity's agentic browser — free download; Comet Plus is a $5 journalism add-on.","url":"https://www.perplexity.ai/comet","category":"productivity","pricing":"freemium","price_note":"browser free worldwide; Comet Plus $5/mo standalone (included with Perplexity Pro $20/mo or Max $200/mo); Comet Agent quotas follow Pro/Max on the hub pricing table","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"A separate product from Perplexity search: Chromium with an assistant that can click and fill. The browser itself went free. Comet Plus is publisher access at $5, not the browser license. Agent depth still tracks the Perplexity seat. This pass hit Cloudflare/timeouts on the Comet URLs — receipts are the Plus announcement and hub pricing, not a live install. Not Dia.","limitations":["perplexity.ai/comet timed out for this bot; confirm download in a browser.","Comet Agent is quota-gated on Free vs Pro/Max — the free browser is not unlimited agent.","Comet Plus is journalism access, not more agent compute.","This is not a second Perplexity search card; citations still live on perplexity.ai."],"strengths":["Chromium-based assistant browser that can click and fill forms.","The browser itself is free; Comet Plus is a cheap $5 journalism add-on.","Agent quotas follow the existing Perplexity Pro/Max seat structure."],"use_for":["Users who want an AI assistant baked into the browser rather than a separate tab.","Existing Perplexity subscribers who can reuse Pro/Max agent quotas."],"skip_when":["You need a guaranteed live download; the Comet URLs timed out during verification.","You expect the free browser to include unlimited agent actions.","You want a standalone search product rather than a browser assistant."],"receipts":["https://www.perplexity.ai/hub/blog/introducing-comet-plus","https://www.perplexity.ai/hub/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"consensus","name":"Consensus","tagline":"Scientific search engine that answers questions with evidence from papers.","url":"https://consensus.app","category":"data","pricing":"freemium","price_note":"free tier; Pro $20/mo ($144/yr), Deep $65/mo ($540/yr)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Useful when you need a quick read on what the research actually says: the consensus meter and study-quality flags keep the answer grounded. It is narrower than Elicit — it answers questions, it does not run a full literature review.","limitations":["Only as good as its paper index.","Summaries flatten nuance and caveats.","Limited to research-question use cases."],"strengths":["Consensus meter and study-quality flags keep answers grounded in papers.","Quick read on what the research actually says about a question."],"use_for":["Getting a fast, evidence-backed answer to a research question.","Checking the balance of evidence before diving into primary sources."],"skip_when":["You need a full literature review rather than a focused answer.","The nuance and caveats of a study matter more than the headline.","Your topic sits outside the indexed paper coverage."],"receipts":["https://consensus.app","https://help.consensus.app/en/articles/10087865-subscription-plans"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"context7-mcp","name":"Context7","tagline":"Upstash Context7 — current library docs in the agent's context, via hosted MCP.","url":"https://context7.com","category":"mcp","pricing":"freemium","price_note":"Free $0 (1,000 API calls/mo, public repos); Pro $10/seat/mo (5,000/seat); hosted https://mcp.context7.com/mcp","free_tier":true,"open_source":true,"api":true,"self_host":false,"verdict":"ship","verdict_text":"The docs MCP people actually keep on: resolve a library, pull current examples, stop hallucinating last year's API. Hosted at mcp.context7.com. Free is 1,000 calls/month; it used to be much higher. Queries go to Upstash.","limitations":["Free tier is 1,000 API calls/month (cut hard in Jan 2026, now 1,000 + 20/day bonus after cap).","All queries route through Upstash; library name and query text leave the machine.","Pro is $10/seat/mo then $10 per 1,000 extra calls."],"strengths":["Current, version-specific library docs instead of training-cut snapshots.","Hosted Streamable HTTP plus a local npm/stdio option; npx ctx7 setup.","MIT server, OAuth or API key, works across Cursor / Claude / VS Code."],"use_for":["Coding agents that keep inventing APIs the package no longer has.","Public-repo library work on the free 1,000-call month."],"skip_when":["1,000 free calls/month will not cover a \"always use Context7\" rule.","The repo cannot leave your machine; there is no offline mode.","You need private-repo parsing — that is Pro, billed extra."],"receipts":["https://github.com/upstash/context7","https://context7.com/plans","https://mcp.context7.com/mcp"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":false},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"continue-dev","name":"Continue","tagline":"Open-source IDE agent (VS Code/JetBrains) with hub configs, BYOK, and local models.","url":"https://continue.dev","category":"coding","pricing":"freemium","price_note":"OSS extension free; Continue Hub has paid team features","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Noemium has not field-run Continue in this wave — listing is from the project site and GitHub.","Hub vs local config is two products; do not assume the OSS extension equals the cloud hub."],"receipts":["https://continue.dev","https://github.com/continuedev/continue"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"copy-ai","name":"Copy.ai","tagline":"Marketing copy generator with workflow automation for go-to-market teams.","url":"https://www.copy.ai","category":"writing","pricing":"freemium","price_note":"free trial (2k words/mo); paid plans from $1,000/mo (Growth)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Strong for short-form marketing copy and sales emails. The workflow builder is useful for repetitive GTM tasks, but long-form content still needs heavy editing and brand voice tuning.","limitations":["Long-form drafts often feel formulaic.","Brand voice customization requires paid plans.","Output quality varies across content types."],"strengths":["Strong short-form marketing copy and sales email generation.","Workflow builder automates repetitive GTM tasks."],"use_for":["Short marketing copy and sales emails at scale.","Repetitive go-to-market workflows that need templated output."],"skip_when":["You need long-form drafts without heavy editing.","Brand voice customization is required on a tight budget.","Output consistency across every content type is critical."],"receipts":["https://www.copy.ai","https://www.copy.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"cosyvoice","name":"CosyVoice","tagline":"Text-to-speech and voice-cloning model family from Alibaba Cloud.","url":"https://github.com/QwenAudio/CosyVoice","category":"audio","pricing":"paid","price_note":"CosyVoice 2 from ~$0.287/10,000 characters; CosyVoice 3.5 Plus from ~$0.22/10,000 characters; CosyVoice voice creation free; Qwen-TTS voice creation $0.20/voice; free quota: 10 voice creations in first 90 days (Singapore region)","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"locked","verdict":"situational","verdict_text":"Best-in-class Chinese TTS naturalness, plus open-source weights you can self-host. Pricing is token-based and can be confusing compared with per-minute audio APIs; English quality is good but not market-leading.","limitations":["Token/character pricing can be confusing vs per-minute audio APIs.","Commercial rights and voice-cloning ethics need review.","English and non-Chinese quality are good but not market-leading.","Same Alibaba Cloud region/compliance friction as other Qwen services."],"strengths":["Top-tier naturalness for Chinese text-to-speech.","Open-source weights that can be self-hosted.","API and voice-creation options on Alibaba Cloud."],"use_for":["Chinese-language applications that need natural-sounding TTS.","Teams that can self-host to keep voice data off shared cloud endpoints."],"skip_when":["You need simple per-minute pricing and predictable audio costs.","English or non-Chinese output must be market-leading.","Alibaba Cloud region and compliance requirements are a blocker."],"receipts":["https://github.com/QwenAudio/CosyVoice","https://www.alibabacloud.com/help/en/model-studio/voice-design-user-guide","https://cloudprice.net/models/alibaba-cosyvoice-2","https://cloudprice.net/models/alibaba-cosyvoice-3-5-plus","https://guijiagi.com/posts/ai-tts-2026-comparison/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"crewai","name":"CrewAI","tagline":"Python framework for multi-agent crews — roles, tasks and processes.","url":"https://www.crewai.com","category":"agents","pricing":"freemium","price_note":"open source (MIT); cloud Free Basic (50 exec/mo); Enterprise custom","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"ship","verdict_text":"Roles, tasks, and processes map cleanly onto real teams, and the examples get a prototype running fast. Good for demos and early workflows. Production means debugging delegation chains and watching token costs climb.","limitations":["Multi-agent loops burn tokens fast.","Debugging delegation chains between agents is painful.","Production observability needs external tooling."],"strengths":["Roles, tasks and processes map onto real teams without inventing a graph DSL.","Examples get a multi-agent prototype running fast."],"use_for":["Demos and early crew workflows in Python.","A first multi-agent sketch before you pay for observability."],"skip_when":["Token burn from agent loops is already a problem.","You need to debug delegation chains in production.","You require first-party production observability."],"receipts":["https://github.com/crewAIInc/crewAI","https://www.crewai.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"crusoe-cloud","name":"Crusoe Cloud","tagline":"AI cloud platform for managed inference, serverless fine-tuning, and self-serve GPU deployments.","url":"https://crusoecloud.com","category":"models-api","pricing":"paid","price_note":"Serverless inference from $0.03–$4.40 per 1M tokens; GPU instances from $2.00–$4.29/GPU-hr; fine-tuning from $0.40–$10.00 per 1M tokens.","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Spot pricing is only suitable for fault-tolerant workloads; guaranteed reserved rates require contacting sales."],"receipts":["https://crusoecloud.com","https://www.crusoe.ai/cloud/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"cursor","name":"Cursor","tagline":"AI-first code editor (VS Code fork) with agent mode and inline edits.","url":"https://cursor.com","category":"coding","pricing":"freemium","price_note":"Hobby free; Pro $20/mo; Pro+ $60/mo (3× Agent limits); Ultra $200/mo (20×); Teams from $40/user/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["claude-sonnet-5"],"model_routing":"bundled","verdict":"ship","verdict_text":"Still the default AI editor, and for a reason: agent mode and tab completions genuinely save hours. The model routing is a black box — you never quite know what you're talking to — but on Sonnet-class models the output is consistently good.","limitations":["Pro plan token limits throttle heavy agent sessions mid-month.","Closed source; no way to self-host or audit data handling.","Occasional context-window degradation on very large repos."],"strengths":["Agent mode and tab completions that actually save hours on everyday edits.","Default AI editor for a reason: Sonnet-class output is consistently good.","Fast to start if you already live in a VS Code layout."],"use_for":["Daily product coding where you want an IDE, not a terminal ritual.","Multi-file agent runs when you can sit with the diff."],"skip_when":["You need to audit or self-host the editor.","Heavy agent months where Pro token limits cut you off mid-sprint.","You refuse black-box model routing."],"receipts":["https://cursor.com/pricing","https://cursor.com/docs"],"affiliate":"none","evidence_tier":"source-verified","guide":{"install":"Download from cursor.com/download. Sign in. Optional: cursor-agent CLI for headless runs.","requirements":["A Cursor account","A project folder"],"cost":"Hobby $0, Pro $20/mo, Pro+ $60/mo, Ultra $200/mo, Teams from $40/user/mo — as of the catalog card.","security":"Closed editor. Trains-on-inputs is yes on the card. Privacy mode is a setting, not a proof. Field guide: /agents/cursor/.","breaks_when":["A heavy agent month hits Pro token limits mid-sprint.","You need to audit or self-host the editor."]},"momentum":"blueshift","featured":true,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"dagger","name":"Dagger","tagline":"Programmable CI platform that replaces artisanal scripts with a portable DAG execution engine and language SDKs.","url":"https://dagger.io","category":"dev-infra","pricing":"freemium","price_note":"Individual plan is free with 1 user and 1,000,000 events per month; Team is $50 per month after a 2-week trial; Enterprise is custom.","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Individual plan is limited to one user, 1,000,000 monthly events, and one month of run history."],"receipts":["https://dagger.io","https://docs.dagger.io","https://github.com/dagger/dagger","https://dagger.io/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"daisyui","name":"DaisyUI","tagline":"Tailwind CSS component plugin with semantic class names.","url":"https://daisyui.com","category":"design","pricing":"free","price_note":"MIT open source","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Speeds up prototyping by giving ready-made component classes like btn and card. Great for quick admin tools and MVPs, but the look is generic without heavy customization.","limitations":["Visual style is generic out of the box.","Less flexible than utility-first Tailwind for bespoke designs.","Theming requires understanding of the plugin system."],"strengths":["Ready-made component classes like btn and card speed up prototyping.","Semantic class names that are easy to remember.","MIT open source."],"use_for":["Quick admin tools and MVPs built with Tailwind.","Teams that want pre-built components without leaving Tailwind."],"skip_when":["Your design is bespoke and needs utility-first flexibility.","The generic out-of-the-box look is unacceptable.","The team does not understand the plugin theming system."],"receipts":["https://daisyui.com","https://github.com/saadeghi/daisyui"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"daytona","name":"Daytona","tagline":"Secure, elastic infrastructure for running AI-generated code in isolated, stateful sandboxes.","url":"https://www.daytona.io","category":"dev-infra","pricing":"freemium","price_note":"Pay-as-you-go compute with $200 in free credits at sign-up; vCPU $0.0504/h, memory $0.0162/h, storage $0.000108/h after first 5 GiB; startup credits up to $50,000.","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Pricing is usage-based per vCPU, memory, storage, and GPU hours; Windows sandboxes add an OS surcharge and long-running workloads accumulate cost."],"receipts":["https://www.daytona.io","https://www.daytona.io/docs","https://www.daytona.io/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"deepgram","name":"Deepgram","tagline":"Speech-to-text API built for developers — fast, cheap, real-time.","url":"https://deepgram.com","category":"audio","pricing":"freemium","price_note":"$200 free credits; Nova-3 STT from $0.0048/min (promo)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"A solid workhorse STT API: real-time streaming, decent accuracy, and cheaper than most rivals at volume. If self-hosting Whisper adds too much latency, this is the obvious fallback. Accuracy drops on heavy accents and noisy audio, and diarization needs per-case tuning. It's another vendor in your stack, but it does the job.","limitations":["Accuracy on heavy accents and noise below a well-tuned Whisper setup.","Diarization and formatting need tuning per use case.","Another vendor dependency in your voice pipeline."],"strengths":["Workhorse streaming STT: fast, cheap at volume, good enough accuracy.","The obvious hosted fallback when self-hosted Whisper is too slow."],"use_for":["Real-time transcription in a product voice pipeline.","Batch STT where per-minute cost matters more than perfect diarization."],"skip_when":["Heavy accents and noise need a tuned Whisper setup.","Speaker diarization has to be right without per-case work.","You refuse another vendor in the voice stack."],"receipts":["https://deepgram.com","https://deepgram.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"deepinfra","name":"DeepInfra","tagline":"Serverless inference cloud for 100+ open-source LLMs, embeddings, image, video, and speech models.","url":"https://deepinfra.com","category":"models-api","pricing":"paid","price_note":"Pay per token with model-specific rates shown on model cards; private deployments on A100/H100/H200/B200/B300 billed per GPU-hour.","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","No free tier is shown; private deployments and GPU clusters move you from per-token to hourly GPU billing."],"receipts":["https://deepinfra.com","https://deepinfra.com/docs"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"deepseek-api","name":"DeepSeek API","tagline":"Open-weights lab with frontier-ish reasoning at disruptive prices.","url":"https://api-docs.deepseek.com","category":"models-api","pricing":"paid","price_note":"no free API tier; V4 Flash peak $0.44 input (cache miss) / $0.014 cache hit / $1.32 output per 1M, V4 Pro peak $1.32 input / $0.044 cache hit / $3.96 output per 1M; off-peak 50% off (01:00–04:00, 06:00–10:00 UTC); effective 2026-08-16 16:00 UTC","free_tier":false,"open_source":false,"api":true,"self_host":false,"models_used":["deepseek-v4-flash","deepseek-v4-pro"],"model_routing":"locked","verdict":"situational","verdict_text":"The best price-performance around for reasoning workloads — R1-class output at a fraction of frontier cost. The catch is China hosting: plenty of companies simply can't send data there. When compliance allows, it's a no-brainer.","limitations":["Data residency is a compliance blocker for many orgs.","API capacity and reliability have wobbled under demand.","Censorship baked in on some topics.","Complex peak/off-peak rate card with 50% discount outside 01:00–04:00 and 06:00–10:00 UTC from 2026-08-16; cache-hit discounts matter but add opacity."],"strengths":["Best price-performance around for reasoning workloads.","R1-class output at a fraction of frontier API cost.","Off-peak rates are half price, so shifting workloads saves real money."],"use_for":["Reasoning workloads where China-hosted data is compliance-acceptable.","Teams that want frontier-class output without frontier token bills.","Workloads that can run during off-peak windows to cut cost."],"skip_when":["Data residency rules forbid sending workloads to China.","You need predictable capacity and uptime under demand spikes.","Topic censorship or a complex peak/off-peak rate card is a hard no."],"receipts":["https://api-docs.deepseek.com","https://api-docs.deepseek.com/quick_start/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"deepseek-chat","name":"DeepSeek Chat","tagline":"DeepSeek's consumer chat app — not the developer API (that's a separate card).","url":"https://chat.deepseek.com","category":"productivity","pricing":"free","price_note":"web/app chat is free; API is a different product (see DeepSeek API) with peak/off-peak token rates","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["deepseek-v4-flash","deepseek-v4-pro"],"verdict":"situational","verdict_text":"The chat surface for DeepSeek's models. Bot fetches to chat.deepseek.com 403; the app is still the canonical URL. This is not the API rate card. China hosting and topic filters apply. Compare with ChatGPT/Claude on this shelf. No logged-in session in this pass.","limitations":["chat.deepseek.com is aggressive toward bots (403); humans still use the app.","Data residency is a compliance blocker for many orgs (same as the API card).","Censorship on some topics is baked in.","Do not confuse free chat with the paid, peak/off-peak API."],"strengths":["Free consumer chat for DeepSeek's own models.","Direct alternative shelf to ChatGPT/Claude."],"use_for":["Personal chat when you want DeepSeek's models without API setup.","Comparing DeepSeek output against ChatGPT/Claude."],"skip_when":["Data residency or China hosting is a compliance blocker.","You need uncensored responses.","You actually need the API rate card and peak/off-peak pricing."],"receipts":["https://api-docs.deepseek.com","https://www.deepseek.com"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"descript","name":"Descript","tagline":"AI audio and video editor that turns transcripts into a doc-style timeline — record, edit, overdub, and repurpose.","url":"https://www.descript.com","category":"video","pricing":"freemium","price_note":"Free $0; Hobbyist $16/mo (annual) or $24/mo; Creator $24/$35; Business $50/$65; Enterprise custom.","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Descript shines for podcasters, marketers, and solo creators who want to edit audio and video by editing text. Overdub, Studio Sound, and AI clip generation save hours on repetitive post-production. The trade-off is a proprietary cloud project format and usage quotas; heavy teams will outgrow the cheaper tiers quickly.","limitations":["Free plan is capped at 60 minutes of imported/recorded media, 5 GB storage, 720p export, and one-time AI credits.","AI features consume credits that can run out mid-project, and overage costs add up for high-volume creators.","Doc-style editing sacrifices the fine timeline control that traditional editors expect for complex multicam work."],"strengths":["AI audio/video editing through transcript edits, aimed at podcasters and marketers.","Overdub, Studio Sound, and AI clip generation save hours on repetitive post-production.","Doc-style timeline is friendlier than a traditional NLE for non-editors."],"use_for":["Podcast and solo-creator edits where cutting by transcript is faster than a timeline.","Repurposing long recordings into clips, overdub fixes, and marketing assets.","Marketing teams producing lots of short-form content from one source recording."],"skip_when":["Complex multicam work needs frame-accurate timeline control.","Heavy production teams will blow through quota and AI credits quickly.","You need a non-proprietary, offline project format."],"receipts":["https://www.descript.com","https://www.descript.com/pricing","https://help.descript.com/hc/en-us/articles/27653385141517-Understanding-Your-Legacy-and-Sunset-Plan"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"devin","name":"Devin","tagline":"Autonomous cloud AI software engineer that plans, codes, tests, and opens PRs; now bundled with the Devin Desktop IDE.","url":"https://devin.ai","category":"coding","pricing":"freemium","price_note":"Free $0/mo (limited agent quota, unlimited inline edits); Pro $20/mo adds Devin Cloud and frontier models; Max $200/mo; Teams from $80/mo + $40/full dev seat; the old $500/mo Team entry was retired in April 2026.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Cloud agent that takes tickets from prompt to tested PR in its own VM (browser, terminal, git), working async. Fits teams with well-scoped, verifiable tickets and a strong review culture; not a fit for unstructured vibe coding. Main friction is metered usage that can spike on large tasks, plus the permission and security overhead of keeping an autonomous agent away from production secrets.","limitations":["Free tier excludes autonomous Devin Cloud sessions and frontier models.","Quota/ACU-based usage can incur extra charges at API pricing on complex tasks.","No self-serve self-host (Enterprise custom only); autonomous runs still need human review and RBAC guardrails."],"strengths":["Cloud agent takes tickets from prompt to tested PR in its own VM.","Async workflow with browser, terminal, and git access fits remote tasks."],"use_for":["Teams with well-scoped, verifiable tickets and a strong review culture.","Tasks where an autonomous agent can work while humans are offline."],"skip_when":["You need a free autonomous Cloud session or frontier models on the free tier.","Usage quotas or ACU charges would spike on large, complex tasks.","You cannot accept the security overhead of an autonomous agent near production secrets."],"receipts":["https://devin.ai/pricing","https://docs.devin.ai/api-reference/overview","https://cognition.com/blog/new-self-serve-plans-for-devin"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"dia","name":"Dia","tagline":"The Browser Company's AI browser — free Chromium shell, paid Answers/Days; macOS Apple Silicon.","url":"https://www.diabrowser.com","category":"productivity","pricing":"freemium","price_note":"Better Browser free; Better Answers $20/mo; Better Days $100/mo; homepage includes a 14-day Better Days trial; macOS 14+ M1+","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Arc's successor as an AI browser, now with a public three-rung price table. Free is a real browser (profiles, split view, ad/tracker block, sync). $20 is page-chat; $100 is Morning Brief / decks / meeting prep. /pricing 404s — /plans is the receipt. Windows is still a waitlist question. Not Comet, not ChatGPT in Chrome.","limitations":["Public download is macOS 14+ Apple Silicon; Windows is not a general GA claim on the homepage.","Usage limits apply on paid tiers; the table does not publish chat quotas as numbers.","Default content-sharing to improve Dia is on unless you turn it off in Settings > Privacy.","Atlassian ownership is a procurement fact, not a feature list."],"strengths":["Free tier is a real browser with profiles, split view, ad/tracker block and sync.","Clear three-rung pricing table for page-chat and premium AI features.","Top tier adds Morning Brief, decks and meeting prep inside the browser."],"use_for":["macOS 14+ Apple Silicon users who want an AI-native browser replacement.","Teams that need page-chat and meeting-prep features inside the browser."],"skip_when":["You are on Windows or Intel Macs and need a general-availability download today.","You need published chat quotas for paid tiers before buying.","Default content-sharing toggles must stay off and manual opt-out is not acceptable."],"receipts":["https://www.diabrowser.com/plans","https://www.diabrowser.com/privacy"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"dify","name":"Dify","tagline":"Source-available LLM app platform with agent workflows and RAG.","url":"https://dify.ai","category":"agents","pricing":"freemium","price_note":"Cloud Sandbox free; self-host Community free under restricted license; cloud Professional/Team paid yearly","free_tier":true,"open_source":false,"api":true,"self_host":true,"verdict":"ship","verdict_text":"The fastest way to build a RAG chatbot or agent workflow without writing a backend. Self-hosting keeps data in-house, but the UI can feel rigid for complex custom logic.","limitations":["The Dify license restricts multi-tenant use and frontend rebranding, so it is not OSI open source.","Complex branching logic can outgrow the visual editor.","Self-hosted setup needs ops effort for scaling.","Some enterprise SSO features require cloud plans."],"strengths":["Fastest path to a RAG chatbot or agent workflow without writing a backend.","Self-host keeps the data in-house when the license fits."],"use_for":["Internal RAG apps and agent demos that should ship this week.","Teams that want a visual builder plus an optional self-host."],"skip_when":["You need OSI open source or multi-tenant rebranding the license forbids.","Branching logic has already outgrown a visual editor.","Enterprise SSO is required and you are not on a cloud plan."],"receipts":["https://github.com/langgenius/dify","https://github.com/langgenius/dify/blob/main/LICENSE"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"doberman","name":"Doberman","tagline":"Runtime guard dog for AI coding agents that blocks unsafe tool calls before they execute.","url":"https://github.com/fu351/Doberman-Core","category":"agents","pricing":"free","price_note":"Apache-2.0 open source core; enterprise platform on roadmap","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"Sits on the MCP execution path and gives every tool call a PASS, AUTH, or BLOCK verdict, stopping destructive commands, secret exfiltration, and prompt-injection-driven actions. It fails closed and logs redacted decisions. The project is early but already ships CI-proven parity matrices and a plugin API for custom guardrails.","limitations":["Some host integrations are still experimental or manual to wire.","Adaptive behavioral features currently run only on the MCP proxy path, not all host hooks.","Static egress parsing can be evaded by runtime-constructed URLs or non-verb channels."],"strengths":["PASS/AUTH/BLOCK verdict on every tool call before it executes.","Fail-closed with redacted decision logs.","CI-proven parity matrices and a plugin API for custom guardrails."],"use_for":["Securing MCP-based coding agents against destructive or exfiltrating calls.","Teams that need an open, auditable guardrail they can self-host."],"skip_when":["Your host integration is still experimental or manual to wire.","You need behavioral guards beyond the MCP proxy path.","Runtime-constructed URLs or non-verb channels would evade static egress parsing."],"receipts":["https://github.com/fu351/Doberman-Core","https://github.com/fu351/Doberman-Core/blob/main/docs/ADAPTER_GUIDE.md"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"dspy","name":"DSPy","tagline":"Python framework that compiles declarative LLM programs with optimizers instead of hand-written prompts.","url":"https://dspy.ai","category":"data","pricing":"free","free_tier":true,"open_source":true,"api":false,"self_host":true,"model_routing":"byok","limitations":["Radar: not field-run in this catalog wave.","Requires your own model provider keys; optimization quality depends on labeled examples and metrics."],"receipts":["https://dspy.ai"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"dust","name":"Dust","tagline":"Team agents over company data: connectors, workflows, and a managed workspace.","url":"https://dust.tt","category":"automation","pricing":"freemium","free_tier":true,"open_source":true,"api":true,"self_host":false,"limitations":["Radar: connector security not reviewed here.","A workspace product, not a drop-in n8n replacement."],"receipts":["https://dust.tt","https://github.com/dust-tt/dust"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"e2b","name":"E2B","tagline":"Isolated cloud sandboxes for agents — firecracker VMs billed per second of CPU/RAM.","url":"https://e2b.dev","category":"agents","pricing":"freemium","price_note":"Hobby $0 + usage (one-time $100 credits, 1h sessions, 20 concurrent); Pro $150/mo + usage (24h sessions, 100 concurrent); default 2 vCPU $0.000028/s","free_tier":true,"open_source":true,"api":true,"self_host":false,"verdict":"situational","verdict_text":"The sandbox other agent products quietly run on. Hobby is usage plus a one-time $100 credit pile; Pro is $150 plus the same per-second meters. You still write the agent. Not Browser Use (that's a browser); this is a VM. No production agent parked here.","limitations":["Seat fee and usage are separate; idle sandboxes still tick CPU/RAM.","Hobby session cap is 1 hour; long jobs need Pro.","Self-host of the commercial control plane is not the public SKU.","Concurrency above 100 is a Pro extra purchase, not included."],"strengths":["Isolated Firecracker VM sandboxes billed per second of CPU/RAM.","Generous Hobby tier with $100 one-time credits and 20 concurrent sessions.","The runtime several agent products quietly rely on."],"use_for":["Running untrusted agent code in isolated cloud VMs.","Prototyping agents that need a real Linux environment, not a browser."],"skip_when":["You need sessions longer than the 1-hour Hobby cap without paying Pro.","The budget cannot cover both seat fee and per-second usage.","You want to self-host the commercial control plane."],"receipts":["https://e2b.dev/pricing","https://github.com/e2b-dev/E2B"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"elevenlabs","name":"ElevenLabs","tagline":"The reference standard for AI voice — TTS, cloning, dubbing and voice agents.","url":"https://elevenlabs.io","category":"audio","pricing":"freemium","price_note":"free tier (10k credits/mo, non-commercial); Starter $6/mo, Creator $22/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["eleven-v3"],"verdict":"ship","verdict_text":"Still the reference for AI voice quality, and the API is solid enough to build real products on. Voice cloning and multilingual dubbing mostly just work. The credit system is confusing, the best models burn through credits fast, and Creator doubled in price in 2026. Worth it if you need the best voice, but budget carefully.","limitations":["Credit-based pricing makes budgeting confusing.","Best voices and models burn credits fast.","Cloning consent flows add friction even for legitimate use cases.","2026 repricing — Creator went from $11 to $22 with a bigger quota."],"strengths":["Still the reference for AI voice quality, with an API you can ship on.","Cloning and multilingual dubbing mostly just work."],"use_for":["Product voice, dubbing, or agents where quality is the bottleneck.","A first clone before you commit a production credit budget."],"skip_when":["You cannot forecast credit burn on the best models.","Creator-tier repricing already broke the budget.","Cloning consent flows are a deal-breaker even for legitimate use."],"receipts":["https://elevenlabs.io","https://elevenlabs.io/pricing","https://elevenlabs.io/docs"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":true,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"elicit","name":"Elicit","tagline":"AI research assistant that extracts structured data from academic papers.","url":"https://elicit.com","category":"data","pricing":"freemium","price_note":"free tier; Plus $11/mo, Pro $39/mo (billed yearly)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Saves real time on literature reviews: it pulls methods and results into tables so you do not have to skim dozens of PDFs. Pay for it only if academic papers are a real input to your work.","limitations":["Coverage limited to indexed papers; misses preprints and paywalled work.","Extractions need manual verification.","Pricing tiers have changed repeatedly."],"strengths":["Extracts methods and results from academic papers into structured tables.","Saves time on literature reviews by reducing PDF skimming."],"use_for":["Researchers building evidence tables from indexed academic papers.","Literature reviews where manual extraction would mean skimming dozens of PDFs."],"skip_when":["Your sources are preprints, paywalled papers, or work outside indexed coverage.","You cannot manually verify extractions before citing them.","Pricing tier churn makes budget planning impossible."],"receipts":["https://elicit.com","https://elicit.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"ernie-api","name":"Baidu ERNIE API","tagline":"Enterprise AI platform with ERNIE LLMs, image models, and OpenAI-compatible v2 API.","url":"https://cloud.baidu.com/product-s/qianfan_home","category":"models-api","pricing":"paid","price_note":"ERNIE 5.1 ≤32k ~$0.56/$2.54 per 1M, 32k-128k ~$0.85/$3.10; ERNIE 4.5 Turbo 32K/128K ~$0.11/$0.45; ERNIE X1.1 ~$0.14/$0.56; prices in CNY, USD approximate","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"ERNIE 5.1 ranks high on Chinese leaderboards and the 4.5 Turbo / X1.1 tiers are cheap. The practical path is still mainland Baidu Cloud; international onboarding is harder and pricing is quoted in CNY.","limitations":["Best access path is mainland-China Baidu Cloud; international onboarding is harder.","ERNIE 5.1 and X1.1 are API-only; ERNIE 4.5 weights were released under Apache 2.0 in June 2025, though the hosted Qianfan API remains the practical path for most users.","ERNIE 5.1 pricing is tiered by input length (≤32k vs 32k-128k).","Pure-text reasoning value is weaker than DeepSeek or Qwen on some benchmarks.","ERNIE Lite Pro / Speed Pro are not shown on the current official pricing page."],"strengths":["ERNIE 5.1 ranks high on Chinese leaderboards.","4.5 Turbo and X1.1 tiers are cheap.","OpenAI-compatible v2 API."],"use_for":["Chinese-market applications needing ERNIE models.","Cost-sensitive workloads where 4.5 Turbo/X1.1 pricing wins.","Teams already on Baidu Cloud."],"skip_when":["You need easy international onboarding.","Pure-text reasoning must beat DeepSeek/Qwen.","You need transparent CNY pricing or the omitted Lite Pro/Speed Pro tiers."],"receipts":["https://cloud.baidu.com/product-s/qianfan_home","https://github.com/PaddlePaddle/ERNIE","https://ernie.baidu.com/blog/posts/ernie4.5/","https://cloud.baidu.com/doc/qianfan/s/wmh4sv6ya","https://intl.cloud.baidu.com/en/doc/qianfan/s/Jm8r1826a-intl-en"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"everything-mcp","name":"Everything MCP","tagline":"Official everything MCP — a test server that exercises the protocol, not a tool you run.","url":"https://github.com/modelcontextprotocol/servers/tree/main/src/everything","category":"mcp","pricing":"free","price_note":"MIT reference; npx -y @modelcontextprotocol/server-everything","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"skip","verdict_text":"A protocol gym: prompts, tools, resources, sampling. The official README says it is not intended to be a useful server. Putting it in mcp.json burns context on a test fixture. Keep it off production hosts.","limitations":["Official README: not intended to be a useful server, but rather a test server.","Exercises prompts, resources, sampling — easy to dump a large catalog into context.","Lives in the reference-servers repo; no product, no support desk."],"strengths":["Official reference that hits every MCP primitive in one process.","npx @modelcontextprotocol/server-everything, also a Docker image mcp/everything."],"use_for":["Building or debugging an MCP client, not doing product work.","A conformance check that the host actually speaks prompts and sampling."],"skip_when":["You wanted a kitchen-sink of real tools — this is a test double, not a gateway.","The agent session is billed per token; this server is noisy on purpose.","You copied it from a tutorial into a working mcp.json."],"receipts":["https://github.com/modelcontextprotocol/servers/tree/main/src/everything"],"affiliate":"none","evidence_tier":"source-verified","momentum":"redshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"exa","name":"Exa","tagline":"Neural search engine for finding and ranking web pages by meaning.","url":"https://exa.ai","category":"data","pricing":"freemium","price_note":"free $10 credits/mo; usage-based PayGO; Enterprise custom","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"Far better than keyword search for research and agent context gathering. The API is clean and the results are semantically relevant. Coverage of fresh, niche pages is still uneven.","limitations":["Smaller index than Google/Bing for real-time news.","Pricing can climb quickly at scale.","Some niche or newly published pages are missing."],"strengths":["Semantic search that beats keywords for research and agent context.","Clean API with results ranked by meaning, not just strings."],"use_for":["Agent retrieval that should find the page, not the keyword hit.","Research jobs where semantic ranking matters more than a news index."],"skip_when":["You need Google-scale freshness for breaking news.","Scale would climb the bill faster than a simpler search API.","The pages you need are too new or too niche for the index."],"receipts":["https://exa.ai","https://exa.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"f5-tts","name":"F5-TTS","tagline":"Open-source text-to-speech model using flow matching for fast, faithful speech synthesis.","url":"https://github.com/SWivid/F5-TTS","category":"audio","pricing":"free","free_tier":true,"open_source":true,"api":false,"self_host":true,"model_routing":"bundled","limitations":["Radar: not field-run in this catalog wave.","Pretrained models are licensed CC-BY-NC due to the training dataset, so commercial use is restricted."],"receipts":["https://github.com/SWivid/F5-TTS"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"factory","name":"Factory","tagline":"Autonomous software engineering agents for end-to-end codebase tasks.","url":"https://factory.ai","category":"coding","pricing":"paid","price_note":"custom enterprise pricing","free_tier":false,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Pitches full-codebase agents that plan, write, and review like senior engineers. Impressive demos, but real deployments need tight scoping and human-in-the-loop review. Pricing is enterprise-only.","limitations":["No self-serve tier or transparent pricing.","Requires onboarding and guardrails to avoid large, wrong changes.","Closed platform; limited visibility into model behavior."],"strengths":["Markets full-codebase agents that plan, write, and review code end-to-end.","Impressive demos for autonomous engineering tasks."],"use_for":["Large teams evaluating autonomous codebase agents with vendor support.","Pilots where human-in-the-loop review is already required."],"skip_when":["You need a self-serve tier or transparent pricing.","You cannot invest in onboarding and guardrails to limit wrong changes.","Closed platform and limited model visibility are non-starters."],"receipts":["https://factory.ai","https://docs.factory.ai/harness/agents-md"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"fathom","name":"Fathom","tagline":"AI meeting recorder — unlimited free capture for individuals, paid team CRM/coaching layers.","url":"https://fathom.video","category":"productivity","pricing":"freemium","price_note":"individual Free $0 (unlimited recordings/summaries); Premium $20/mo ($16 billed yearly); Team $19/user/mo ($15 yearly, 2-user min)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"One of the few meeting tools whose free individual tier is actually unlimited recordings, not a three-call tease. Paid is for action items, Ask Fathom, CRM sync, coaching. Bot vs bot-free capture is a preference and a security review. Not a Gong replacement on Free.","limitations":["Free individual omits advanced summaries, action items, and team search.","Team/Business have a 2-user minimum.","CRM field sync and coaching scorecards are Business, not Premium.","Bot-in-meeting vs bot-free (Mac beta) is a policy choice, not a quality guarantee."],"strengths":["Free individual tier has unlimited recordings, not a teaser limit.","Clean split: free for capture, paid for action items/CRM/coaching.","Bot-free capture option on Mac beta."],"use_for":["Individual users who want unlimited meeting recordings.","Small teams upgrading for CRM sync or coaching scorecards.","Mac users who prefer bot-free capture."],"skip_when":["You need action items or team search on the free tier.","Your team is smaller than the 2-user Business minimum.","You expect a full Gong replacement on Free."],"receipts":["https://fathom.video/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"fellow","name":"Fellow","tagline":"Secure AI meeting assistant that records, transcribes, and summarizes meetings with privacy built in.","url":"https://fellow.app","category":"productivity","pricing":"freemium","price_note":"Free plan supports up to 5 users with 5 AI meeting notes and 5 AI recordings per user; Team, Business, and Enterprise plans add automations and CRM integrations.","free_tier":true,"open_source":false,"api":false,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Free plan is limited to 5 users, 5 AI meeting notes per user, and 5 AI recordings per user."],"receipts":["https://fellow.app","https://fellow.app/pricing/"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"no","local_processing":false},"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"fetch-mcp","name":"Fetch MCP","tagline":"Official MCP fetch server — URL to markdown. Can hit local IPs; Python SDK 1.x only.","url":"https://github.com/modelcontextprotocol/servers/tree/main/src/fetch","category":"mcp","pricing":"free","price_note":"MIT reference; uvx mcp-server-fetch or pip mcp-server-fetch","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"The teaching fetch: URL in, markdown out, chunked with start_index. Official README warns it can reach local/internal IPs. Still on MCP Python SDK 1.x; the v2 port is \"in progress\".","limitations":["Official caution: can access local/internal IP addresses.","Requires MCP Python SDK 1.x (`mcp>=1.29.0,<2`); SDK 2 port is unfinished.","Truncates by default (max_length 5000); not a crawler."],"strengths":["Official reference fetch with markdown conversion and chunked reads.","Zero cost, MIT, uvx one-liner."],"use_for":["Agents that need a page as markdown and do not already have a browser MCP.","Demos of stdio MCP without standing up a full browser."],"skip_when":["You cannot accept SSRF to local/internal IPs — the official README flags it.","You are on MCP Python SDK 2.x; this server still requires 1.x.","Playwright or Chrome DevTools MCP already covers the page."],"receipts":["https://github.com/modelcontextprotocol/servers/tree/main/src/fetch"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"figma","name":"Figma","tagline":"Design tool of record; Figma Make (prompt-to-app) is bundled into Full seats, not a separate SKU.","url":"https://www.figma.com","category":"design","pricing":"freemium","price_note":"Starter free (150 AI credits/day, cap 500/mo); Professional Full seat $16/mo billed yearly; Make is Full-seat only on paid plans","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Still the file everyone already has. Make is Figma's AI app builder, sold as Full-seat access plus monthly AI credits (3,000 on Professional), not as Make.com. Starter credits are a taste. Heavy Make use burns the credit pool and then pay-as-you-go. Not a field test of Make output quality.","limitations":["Figma Make is Full-seat on paid plans; Dev/Collab seats are not the Make license.","AI credits are per seat; unused teammate credits do not cover your overage.","Starter AI is 150/day with a 500/mo cap — enough to demo, not to ship.","This is not Make.com (Integromat); automations live on that other card."],"strengths":["Still the design file everyone already has.","Figma Make is bundled into Full seats, not a separate SKU.","Professional tier includes 3,000 monthly AI credits."],"use_for":["Teams already standardized on Figma for design handoff.","Users who want prompt-to-app output inside the familiar Figma surface.","Starter-tier users who only need a small daily AI taste."],"skip_when":["You need Make access without paying for a Full seat.","Heavy Make use would burn through the monthly credit pool.","You are looking for Make.com automations, not Figma's AI builder."],"receipts":["https://www.figma.com/pricing/","https://help.figma.com/hc/en-us/articles/31304485164695-Create-a-Figma-Make-file","https://help.figma.com/hc/en-us/articles/33459875669015-How-AI-credits-work"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"filesystem-mcp","name":"Filesystem MCP","tagline":"Official MCP reference server that exposes a local directory to the agent.","url":"https://github.com/modelcontextprotocol/servers","category":"mcp","pricing":"free","price_note":"MIT reference; npx -y @modelcontextprotocol/server-filesystem <allowed-dir>","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"situational","verdict_text":"The teaching example: an allowlisted folder over stdio. Fine for a demo, and still the snippet every mcp.json tutorial copies. Most coding agents already have files; adding this is extra disk surface, not a new capability.","limitations":["Lives in the reference-servers repo, not a product with a support desk.","stdio process can read whatever you allowlisted — the allowlist is the security model.","Does not replace a coding agent's own file tools."],"strengths":["Official reference from the MCP steering group, not a random registry listing.","Allowlist is a path argument — you can keep it smaller than $HOME.","Zero cost, MIT, npx one-liner."],"use_for":["A first MCP install to prove the client actually talks stdio.","Agents that are not already sitting inside the repo they need to read."],"skip_when":["The agent already has workspace tools (Claude Code, Cursor, Codex, Aider).","You would pass $HOME or / as the allowed path.","You needed a package manager for many servers — that is Pharos, not this."],"receipts":["https://github.com/modelcontextprotocol/servers"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"no","local_processing":true},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"firecrawl","name":"Firecrawl","tagline":"Turn entire websites into clean, LLM-ready markdown or structured data.","url":"https://www.firecrawl.dev","category":"data","pricing":"freemium","price_note":"open source self-host free; cloud 1,000 credits/mo free, then Hobby/Standard/Growth paid","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"ship","verdict_text":"The go-to tool for scraping site content into markdown for RAG. Handles dynamic sites well and the self-hosted option keeps costs down. Very large sites need rate-limiting and chunking discipline.","limitations":["Large sites require careful rate limiting and caching.","Complex JavaScript apps can still break extraction.","Cloud pricing climbs with page volume."],"strengths":["Go-to scrape-to-markdown path for RAG, including many dynamic sites.","Self-host option keeps cloud page-volume pricing off the critical path."],"use_for":["Turning a docs site or blog into LLM-ready markdown.","RAG pipelines that need structured extraction, not raw HTML."],"skip_when":["The target is a huge site you cannot rate-limit and chunk.","The app is a brittle JavaScript SPA that still breaks extraction.","Cloud crawl volume is the only option and the bill would spike."],"receipts":["https://www.firecrawl.dev","https://github.com/mendableai/firecrawl"],"affiliate":"none","evidence_tier":"source-verified","guide":{"install":"Cloud at firecrawl.dev or self-host from mendableai/firecrawl. Start with the cloud free credits.","requirements":["A target site you are allowed to crawl","A sink (markdown files or a vector store)"],"cost":"OSS self-host is free. Cloud: 1,000 credits/mo free, then Hobby/Standard/Growth. Page volume is the bill.","security":"Crawling sends the target HTML to Firecrawl Cloud unless you self-host. Do not point it at authenticated app pages you do not own.","breaks_when":["The site is a brittle SPA and extraction comes back empty.","You crawl a huge site with no rate limit and get banned."]},"momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"firecrawl-mcp","name":"Firecrawl MCP","tagline":"Hosted Firecrawl MCP — scrape, search, parse. Keyless free tier; full tools need a Firecrawl key.","url":"https://docs.firecrawl.dev/mcp-server","category":"mcp","pricing":"freemium","price_note":"hosted https://mcp.firecrawl.dev/v2/mcp; keyless scrape/search/parse (rate-limited); full tools need a Firecrawl key or OAuth. Firecrawl Free $0 = 1,000 credits/mo; Hobby $16/mo billed yearly (https://www.firecrawl.dev/pricing). npx firecrawl-mcp","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"ship","verdict_text":"The web MCP that actually returns markdown: hosted at mcp.firecrawl.dev, keyless scrape/search/parse for a first try, then a Firecrawl key or OAuth for crawl/map/agent. Self-host and npx firecrawl-mcp exist. Credits are the Firecrawl plan, not a second bill.","limitations":["Keyless hosted tools are scrape, search, and parse only, and rate-limited.","Crawl, map, interact, and agent still need a Firecrawl API key or OAuth.","Credits do not roll over on self-serve plans (Firecrawl pricing, 2026-08-26)."],"strengths":["Hosted Streamable HTTP with a keyless free slice (scrape, search, parse).","Same Firecrawl credits as the API — Free is 1,000 credits/month, no extra MCP SKU.","Local npx firecrawl-mcp and self-host when the URL should not leave your cloud."],"use_for":["Coding agents that need a live page as markdown, not a browser MCP.","Keyless first try of scrape/search/parse before creating a Firecrawl key."],"skip_when":["Playwright or Chrome DevTools MCP already covers the page you would scrape.","Crawl, map, or agent tools — those need a Firecrawl key, not the keyless URL.","The target cannot leave your network; then self-host, do not use mcp.firecrawl.dev."],"receipts":["https://docs.firecrawl.dev/mcp-server","https://github.com/firecrawl/firecrawl-mcp-server","https://www.firecrawl.dev/pricing"],"affiliate":"none","evidence_tier":"field-tested","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":false},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"fireflies","name":"Fireflies.ai","tagline":"Meeting bot that records, transcribes and summarizes across platforms.","url":"https://fireflies.ai","category":"productivity","pricing":"freemium","price_note":"free tier (unlimited transcription, 400 min storage); Pro $10/seat/mo (yearly)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"skip","verdict_text":"Lots of features on paper, but the bot still joins your calls and the summaries are no better than cheaper options. For most people, a bot-free tool like Granola is the cleaner choice.","limitations":["Bot visibly joins calls — awkward with external meetings.","Transcription accuracy is average.","Cluttered UI trying to be everything."],"strengths":["Records, transcribes and summarizes meetings across multiple platforms.","Free tier offers unlimited transcription with 400 minutes of storage.","Pro plan adds admin controls and CRM integrations for teams."],"use_for":["Teams that want a meeting bot integrated with their existing call stack.","Organizations that need admin controls and CRM sync for sales calls."],"skip_when":["External meetings where a visible bot joining the call is awkward.","You can get comparable summaries from cheaper or bot-free alternatives.","A cluttered UI trying to do everything slows down your workflow."],"receipts":["https://fireflies.ai","https://fireflies.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"fireworks-ai","name":"Fireworks AI","tagline":"Fast inference platform for open models — serverless and on-demand GPUs.","url":"https://fireworks.ai","category":"dev-infra","pricing":"freemium","price_note":"$1 starter credits; serverless from $0.10/1M tokens; on-demand GPU from $7/hr","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["llama-4-maverick","qwen3-8-max"],"model_routing":"locked","verdict":"situational","verdict_text":"Fast inference, and the tooling around it — function calling, multimodal — is solid. The catch is you're shopping the same open models across Fireworks, Together and Groq. Price-check them each quarter and pick whoever's cheaper.","limitations":["Differentiation vs Together/Groq unclear for standard inference.","Enterprise pricing opaque.","Smaller free tier than rivals.","GPU on-demand prices rise 2026-09-01 (H100 $7 to $8/hr)."],"strengths":["Fast inference with solid function-calling and multimodal tooling.","Serverless and on-demand GPU options for different workloads.","Useful open-model catalog if you already know the model you want."],"use_for":["Open-model inference when you want a serverless endpoint.","Workloads that need dedicated GPU hours instead of per-token billing."],"skip_when":["You need transparent enterprise pricing before signing.","The free tier is too small for your evaluation.","A rival host is cheaper for the same open model this quarter."],"receipts":["https://fireworks.ai","https://fireworks.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"fish-audio","name":"Fish Audio","tagline":"Expressive TTS, cloning, and STT — S2.1 Pro in the web app and a byte-priced API.","url":"https://fish.audio","category":"audio","pricing":"freemium","price_note":"web app free to start (personal use); API pay-as-you-go, no monthly minimum — s2.1-pro $15/M UTF-8 bytes, s2.1-pro-free $0 (fair use, no TTFA/DPA), ASR transcribe-1 $0.36/audio-hour","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Voice platform with an unusually readable API card: $15 per million UTF-8 bytes of input text (~12h speech), plus a $0 s2.1-pro-free string for prototypes. Free web plan is personal-use. Commercial rights and latency SLAs are paid. Not ElevenLabs, not CosyVoice. No production dub in this pass.","limitations":["Free web plan is personal use; monetized/YouTube needs a paid plan (vendor FAQ).","s2.1-pro-free has no TTFA or DPA guarantees — not a production SLA.","API concurrency starts at 5 in-flight until you prepaid ≥$100.","fish.audio/pricing 404'd; the live card is docs.fish.audio models-pricing."],"strengths":["Unusually readable API pricing by UTF-8 byte of input text.","$0 s2.1-pro-free tier for prototypes.","Covers TTS, voice cloning, and STT in one platform."],"use_for":["Projects that need transparent byte-based TTS billing.","Prototyping voice pipelines without a monthly minimum.","Creators who need cloning plus STT alongside TTS."],"skip_when":["You need commercial rights or latency SLAs without paying.","You cannot meet the $100 prepaid threshold for higher API concurrency.","You need a public pricing page rather than the docs card."],"receipts":["https://docs.fish.audio/developer-guide/models-pricing/pricing-and-rate-limits","https://docs.fish.audio/developer-guide/models-pricing/models-overview"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"flowglad","name":"Flowglad","tagline":"YC-backed, open-source payments and billing infrastructure with a zero-webhook, source-of-truth approach.","url":"https://www.flowglad.com","category":"automation","pricing":"freemium","price_note":"free tier; Plus $40/mo (annual billing, save $120); Max $400/mo (annual billing, save $1,200); Enterprise from $1,500/mo","free_tier":true,"open_source":true,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Worth evaluating if you are building an AI/SaaS app and want billing state as a single source of truth without maintaining webhook handlers. The open-source billing engine and SDK are promising, but the product is young, the documented path is the hosted SaaS, and the pricing table is only on the homepage.","limitations":["Young launch; several billing features (seat-based billing, wallets) are still on the roadmap.","Pricing details live only on the homepage, not on a standalone /pricing page.","The open-source repo is available, but the documented onboarding is the hosted Flowglad service.","Plan usage limits (organization credits, spaces, connections, automations) are visible but not defined in detail.","You still need a payment-processor relationship; Flowglad is not itself a processor."],"strengths":["Open-source billing engine and SDK with a source-of-truth data model.","Zero-webhook approach reduces the integration surface.","YC-backed with transparent, if homepage-only, pricing tiers."],"use_for":["AI/SaaS apps that want billing state managed without custom webhook handlers.","Teams evaluating an open, auditable billing engine alongside a hosted option."],"skip_when":["You need mature seat-based billing or wallet features today.","You want a standalone /pricing page rather than a homepage table.","You need Flowglad itself to be the payment processor."],"receipts":["https://www.flowglad.com","https://github.com/flowglad/flowglad","https://www.ycombinator.com/companies/flowglad","https://www.flowglad.com/terms-of-service"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"flux","name":"FLUX (Black Forest Labs)","tagline":"Open-weights image models from the original Stable Diffusion team.","url":"https://bfl.ai","category":"image","pricing":"freemium","price_note":"open weights (FLUX.2 Klein, Apache-2.0); API from ~$0.014/image","free_tier":true,"open_source":true,"api":true,"self_host":true,"models_used":["flux-2-pro"],"verdict":"ship","verdict_text":"The strongest open-weights image model you can self-host: prompt adherence and text rendering are clearly ahead of SDXL, and the API is cheap. First choice if you need weights, fine-tunes, or privacy.","limitations":["FLUX.1 [dev] weights carry a non-commercial license for production use.","The top pro tier is API-only, no weights.","Local runs want a 24GB GPU for comfortable speed."],"strengths":["Strongest open-weights image model you can actually self-host.","Prompt adherence and text rendering sit ahead of SDXL, and the API is cheap."],"use_for":["Image jobs that need weights, fine-tunes, or privacy.","Production APIs where FLUX quality is enough and Midjourney has no API."],"skip_when":["Commercial use of FLUX.1 [dev] would violate the non-commercial license.","You need the top pro tier and therefore have no weights.","You do not have a 24GB GPU for comfortable local runs."],"receipts":["https://github.com/black-forest-labs/flux","https://docs.bfl.ai/flux_2/flux2_overview"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"footer","name":"Footer","tagline":"Curated gallery of website footer design inspiration.","url":"https://footer.design","category":"design","pricing":"free","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"The only dedicated footer inspiration site that is worth bookmarking. Useful when you are stuck on the bottom half of a landing page, but it is still a gallery, not a component store.","limitations":["No code or assets; screenshots only.","Categories are useful but not exhaustive.","Mobile footer examples are underrepresented."],"strengths":["The only dedicated footer inspiration site worth bookmarking.","Useful when you are stuck on the bottom half of a landing page."],"use_for":["Landing-page designers hunting for footer layout references.","Quickly collecting footer patterns before starting a design."],"skip_when":["You need code or assets rather than screenshots.","Mobile footer examples are critical to your project.","The limited categories do not cover your niche."],"receipts":["https://footer.design"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"framer","name":"Framer","tagline":"Design agent for every step from idea to launch.","url":"https://framer.com","category":"design","pricing":"freemium","price_note":"free for basic sites; paid from $15/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"The best choice for designers who want to ship a real site without writing much code. The editor is smooth, the animations are rich, and the hosting just works.","limitations":["Locked into Framer hosting for the easiest path.","Complex dynamic data still needs workarounds.","Pricing climbs quickly for sites with traffic or CMS needs."],"strengths":["Best path from a designed page to a hosted site without much code.","Editor, animation and hosting are one workflow that actually ships."],"use_for":["Marketing sites and launches where the designer owns the last mile.","Teams that want motion-rich pages without a custom front-end."],"skip_when":["You cannot lock hosting to Framer.","The site needs complex dynamic data.","Traffic or CMS needs would jump the plan too fast."],"receipts":["https://framer.com","https://www.framer.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"friendliai","name":"FriendliAI","tagline":"Frontier AI inference cloud with fast model APIs, dedicated endpoints, and container deployments.","url":"https://friendli.ai","category":"models-api","pricing":"paid","price_note":"Model APIs from $0.03–$4.40 per 1M tokens; dedicated endpoints from $2.90–$12.00/GPU-hr; container pricing is contact sales.","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Container and enterprise deployments are not self-serve and require sales engagement."],"receipts":["https://friendli.ai","https://friendli.ai/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"gemini","name":"Gemini","tagline":"Google's consumer AI assistant — chat, Deep Research, image/video tools, bundled Google AI plans.","url":"https://gemini.google.com","category":"productivity","pricing":"freemium","price_note":"free with Google Account; AI Plus $4.99/mo; AI Pro $19.99/mo; AI Ultra from $99.99/mo (5x) or $199.99/mo (20x)","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["gemini-3-1-pro-preview","gemini-3-7-flash"],"verdict":"situational","verdict_text":"The Gemini app is the assistant, not AI Studio and not Antigravity. Free is usable; Plus is $4.99 for 2× caps; Pro at $19.99 also unlocks Jules and Antigravity entry limits. Caps refresh on a 5-hour window until a weekly ceiling. Sit it next to ChatGPT/Claude, not Grammarly.","limitations":["Usage is compute-weighted, not a simple message count; weekly caps still bite.","Image/video (Nano Banana, Veo, Flow credits) are plan-gated and regional.","This card is the consumer app — Gemini API lives on the Google AI Studio card.","Workspace/enterprise Gemini is a different SKU (Gemini Enterprise)."],"strengths":["Free tier is genuinely usable with just a Google account.","Bundled Google AI plans add chat, Deep Research, image/video tools, and assistant features.","Pro tier unlocks Jules and Antigravity entry limits inside the same subscription.","Plus tier is cheap at $4.99 for 2× caps."],"use_for":["Existing Google users who want an all-in-one AI assistant.","Quick research, image/video tasks, and chat within the Google ecosystem.","Casual users who can stay within weekly caps."],"skip_when":["You need the Gemini API or AI Studio (those are separate cards).","Usage predictability matters; caps refresh on a 5-hour window with a weekly ceiling.","You need Workspace/enterprise admin controls (that is Gemini Enterprise).","Image/video tools or Veo are region-gated for your location."],"receipts":["https://gemini.google/subscriptions/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"gemini-code-assist","name":"Gemini Code Assist","tagline":"Google's AI coding assistant for IDEs, terminal, and Google Cloud workloads.","url":"https://codeassist.google.com","category":"coding","pricing":"paid","price_note":"Standard $19/seat/mo (annual) or $22.80 monthly; Enterprise $45/seat/mo (annual) or $54 monthly","free_tier":false,"open_source":false,"api":false,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Strong contextual suggestions inside VS Code and JetBrains, with deep Google Cloud integration. It lags Cursor and Claude Code on agentic workflows, and enterprise rollout can be slow.","limitations":["Agent capabilities are narrower than Cursor or Claude Code.","Enterprise billing and data handling need Google workspace alignment.","Model updates roll out unevenly across IDE plugins."],"strengths":["Strong contextual suggestions inside VS Code and JetBrains.","Deep Google Cloud integration for cloud-first teams.","Standard seat is cheaper than Cursor/Claude Code-style per-seat tools."],"use_for":["Teams already committed to Google Cloud and Google Workspace.","Developers wanting IDE-native suggestions without switching editors."],"skip_when":["Multi-step agentic workflows are the main job.","Enterprise procurement cannot wait for Google workspace alignment.","You need consistent model updates across every IDE plugin immediately."],"receipts":["https://codeassist.google.com","https://codeassist.google/products/business"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"gemini-nano-banana","name":"Gemini Nano Banana 2","tagline":"Gemini's native image model (gemini-3.1-flash-image) — generate/edit in the Gemini app and Gemini API.","url":"https://ai.google.dev/gemini-api/docs/image-generation","category":"image","pricing":"freemium","price_note":"Gemini app: included with plan caps (Free limited; Plus/Pro/Ultra expand); API billed as Gemini image tokens — confirm live rates on Google AI pricing (not Imagen 4)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Nano Banana is Gemini's image stack, not Imagen 4. Nano Banana 2 is the 3.1 Flash Image workhorse (4K, references, text in image); Lite is the cheap/fast sibling. App access follows Google AI Plus/Pro/Ultra. API is token-priced per resolution. No studio run in this pass.","limitations":["Not Imagen 4 — different model family, different console, different bill.","App image caps are plan- and region-gated (Nano Banana Pro in Search is a separate note).","Lite is not built for multi-reference or long sequential edits.","Gemini API prices move; the image-generation doc is the receipt, not a screenshot of $/image."],"strengths":["Native 4K image generation and editing inside the Gemini app and API.","Supports references and rendered text in images via Nano Banana 2.","Lite variant offers a cheaper, faster sibling for less demanding work.","API bills as Gemini image tokens tied to output resolution."],"use_for":["Gemini app users on Plus/Pro/Ultra who want native image generation.","Developers already calling the Gemini API for image edits and creation."],"skip_when":["You need Imagen 4 capabilities or pricing — this is a different model family.","You want a fixed per-image rate card instead of token-priced API billing.","Lite must handle multi-reference images or long sequential edits."],"receipts":["https://ai.google.dev/gemini-api/docs/image-generation","https://gemini.google/subscriptions/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"git-mcp","name":"Git MCP","tagline":"Official MCP git server — status, diff, log, and it can commit. Python SDK 1.x only.","url":"https://github.com/modelcontextprotocol/servers/tree/main/src/git","category":"mcp","pricing":"free","price_note":"MIT reference; uvx mcp-server-git or pip mcp-server-git","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"situational","verdict_text":"The teaching git MCP: status, diff, log, branch, and also add / commit / checkout / reset. Official README still calls it early development and pins MCP Python SDK 1.x. Most coding agents already have git; this is extra write surface, not a missing capability.","limitations":["Requires MCP Python SDK 1.x (`mcp>=1.29.0,<2`); SDK 2 port is unfinished.","git_commit, git_add, git_reset, git_checkout are write tools on the repo you pass.","README: currently in early development; tools can change."],"strengths":["Official reference from the MCP steering group, uvx one-liner.","Structured git_status / git_diff / git_log without scraping terminal output.","Docker image mcp/git with a bind mount if you do not want uvx."],"use_for":["A first MCP install to prove the client can talk to a local repo.","Agents that are not already sitting inside the git repo they need to read."],"skip_when":["The agent already has git (Claude Code, Cursor, Codex, Aider).","You cannot accept git_commit / git_checkout from an MCP tool.","You are on MCP Python SDK 2.x; this server still requires 1.x."],"receipts":["https://github.com/modelcontextprotocol/servers/tree/main/src/git"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"no","local_processing":true},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"github-copilot","name":"GitHub Copilot","tagline":"Completions, chat, agents, and Copilot CLI — one GitHub Copilot seat, not a third CLI product.","url":"https://github.com/features/copilot","category":"coding","pricing":"freemium","price_note":"Free $0 (2,000 completions/mo, 50 chat requests, CLI, limited agent mode); Pro $10/mo ($15 credits); Pro+ $39/mo ($70 credits); Max $100/mo ($200 credits)","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["gpt-5-6-terra","claude-sonnet-5"],"model_routing":"bundled","verdict":"ship","verdict_text":"The zero-friction option if your team already lives on GitHub. Fast completions, real PR integration, a free tier that's actually usable, and Copilot CLI is on that same seat (not a separate catalog card). Agent mode still trails Cursor and Claude Code. Pro+ unlocks third-party agents (Claude Code, Codex) inside the GitHub bill.","limitations":["Agent mode still lags Cursor and Claude Code on multi-step tasks.","Output quality varies noticeably between the models in its picker.","Chat, agent, CLI, and cloud agent all draw GitHub AI Credits on paid plans.","Billing moved to AI Credits (2026-06); autocomplete unlimited on paid."],"strengths":["Zero friction if the team already lives on GitHub — one seat, IDE + CLI.","A free tier that is actually usable (completions, chat, limited agent).","PR integration that other agents still treat as an afterthought."],"use_for":["Teams standardized on GitHub who want completions and light agent work.","A first coding agent that should not be a second vendor conversation."],"skip_when":["Multi-step agent work is the job — Cursor and Claude Code still lead.","You need a predictable bill; chat, agent, CLI, and cloud share AI Credits.","You want to self-host or keep code off GitHub's models."],"receipts":["https://github.com/features/copilot/plans","https://docs.github.com/en/copilot"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"github-mcp","name":"GitHub MCP","tagline":"GitHub's official MCP server — repos, issues, PRs, locally or hosted at api.githubcopilot.com.","url":"https://github.com/github/github-mcp-server","category":"mcp","pricing":"free","price_note":"MIT; local Docker/Go or remote https://api.githubcopilot.com/mcp/; PAT or OAuth, no GitHub seat fee for the server","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"ship","verdict_text":"The GitHub-shaped MCP: local image ghcr.io/github/github-mcp-server or the hosted remote. Toolsets and a read-only URL exist. The old npm @modelcontextprotocol/server-github is dead (April 2025) — do not copy it.","limitations":["Remote extra tools (Copilot coding agent) are not in the local binary.","A PAT or OAuth token with repo scope is a repo-scope token sitting in mcp.json.","npm @modelcontextprotocol/server-github is unsupported as of 2025-04."],"strengths":["Official GitHub server with remote hosted option — no local runtime required.","Toolsets and /readonly URLs so you do not dump the whole API into context.","MIT, v1.11.0 (2026-08-25), Docker image ghcr.io/github/github-mcp-server."],"use_for":["Agents that triage issues, read PRs, or search code on GitHub.","Hosts that speak remote MCP and should not run a local GitHub binary."],"skip_when":["You still have @modelcontextprotocol/server-github in mcp.json — that package is retired.","The repo cannot leave GitHub's hosted agent or your PAT scopes.","You needed GitLab or Gitea; this server is GitHub-only."],"receipts":["https://github.com/github/github-mcp-server","https://github.com/github/github-mcp-server/blob/main/docs/remote-server.md"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":false},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"glm-api","name":"GLM API","tagline":"Zhipu AI's OpenAI-compatible API for GLM text, vision, image and video models, plus flat-rate coding plans.","url":"https://open.bigmodel.cn/pricing","category":"models-api","pricing":"paid","price_note":"GLM-5.3 (1M context) approx ¥8/¥28 per 1M and temporarily free as of scan; GLM-5.2 same; GLM-5.1 ¥6–8/¥24–28, GLM-5-Turbo ¥5–7/¥22–26, GLM-4.7-Flash free","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["glm-5-3","glm-5-2"],"model_routing":"locked","verdict":"situational","verdict_text":"GLM-5.2 is a serious open-weights flagship with a huge context window, and the flat-rate Coding Plan is genuinely different. The catch is Zhipu's international surface is thinner than Qwen's, with CNY pricing, lower brand recognition and a promo price that may not last.","limitations":["The hosted API is proprietary; some GLM model weights are available separately for self-hosting.","Pricing page is CNY-only and tiered by input/output length.","GLM-5.3 and GLM-5.2 were temporarily free at scan time — promo pricing may end without notice.","Western brand recognition and documentation are lower than Qwen/DeepSeek/Kimi.","International sign-up and billing can be less straightforward."],"strengths":["GLM-5.2 is a serious open-weights flagship with a huge context window.","Flat-rate Coding Plan is a genuinely different pricing option.","OpenAI-compatible API covers text, vision, image and video models."],"use_for":["Chinese-lab pricing with a flat-rate coding subscription.","GLM-family multimodal calls through an OpenAI-compatible endpoint."],"skip_when":["You need English-first docs and support comparable to Qwen.","CNY-only pricing and tiered rate cards are a compliance headache.","The temporary free promo ending would break your cost model."],"receipts":["https://open.bigmodel.cn/pricing","https://www.aipricing.guru/z-ai-subscription-pricing/","https://therouter.ai/blog/zhipu-glm-api-complete-guide"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"glm-coding-plan","name":"GLM Coding Plan","tagline":"Flat-rate monthly subscription for coding-optimized GLM models via an OpenAI-compatible API.","url":"https://z.ai/subscribe","category":"coding","pricing":"paid","price_note":"Lite ~$18/mo, Pro ~$72/mo, Max ~$160/mo (annual billing $151.20/$604.80/$1,344/year); rate/allowance caps apply","free_tier":false,"open_source":false,"api":true,"self_host":false,"models_used":["glm-5-3","glm-5-2"],"model_routing":"locked","verdict":"situational","verdict_text":"A flat-rate coding-agent subscription is a rare alternative to per-token billing, and the OpenAI-compatible API plugs into Claude Code, Cline or Cursor. The value depends on whether you hit the plan's rate caps; the entry price also jumped from $3 to $18, so watch for further moves.","limitations":["Price rose from introductory $3/mo to $18/mo; may rise again.","Rate and allowance caps apply within flat-rate plans.","Ecosystem smaller than Cursor/GitHub Copilot.","Requires an existing agent tool to be useful; not a standalone IDE."],"strengths":["Flat-rate monthly pricing avoids per-token surprise bills.","OpenAI-compatible API fits existing agent tools like Claude Code or Cline.","Tiered plans from Lite to Max for different coding volumes."],"use_for":["Developers who want predictable monthly costs for a coding agent API.","Teams already using OpenAI-compatible agent tools and need a GLM option."],"skip_when":["You need a standalone IDE rather than an API to plug into one.","Your usage will burst through the published rate or allowance caps.","You want the largest ecosystem of integrations and support."],"receipts":["https://z.ai/subscribe","https://www.aipricing.guru/z-ai-subscription-pricing/","https://blogs.truescho.com/en/glm-coding-plan-zai/","https://bota.chat/z-ai/glm-coding-plan/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"google-ai-studio","name":"Google AI Studio / Gemini API","tagline":"Gemini API with a real free tier — huge context, multimodal, generous limits.","url":"https://aistudio.google.com","category":"models-api","pricing":"freemium","price_note":"free tier (text models); paid from $0.75/1M input (Gemini 3.7 Flash promo till 2027-01)","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["gemini-3-1-pro-preview","gemini-3-7-flash"],"model_routing":"locked","verdict":"ship","verdict_text":"Gemini has the most useful free API tier going, and paid plans undercut rivals for long-context and multimodal work. Million-token context makes document workflows cheap here that get expensive elsewhere. Watch the fast-moving model names and the 2027-01 promo expiry.","limitations":["Free-tier prompts may be used for training — read the terms.","Safety filters occasionally over-trigger.","API surface and model names change quickly.","Image/video generation is paid-tier only; Flash promo prices double 2027-01-01."],"strengths":["Most useful free Gemini API tier going, plus cheap long-context paid plans.","Million-token context makes document jobs cheap that get expensive elsewhere."],"use_for":["Multimodal and long-document prototypes on the free tier.","Production Gemini calls when Flash promo pricing still holds."],"skip_when":["Free-tier prompts used for training is a non-starter.","Safety filters over-trigger on your domain.","You need image or video generation on the free tier."],"receipts":["https://ai.google.dev/gemini-api/docs/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"google-antigravity","name":"Google Antigravity","tagline":"Google's agentic development platform — free individual tier, paid quota via AI Pro/Ultra, orgs via Cloud.","url":"https://antigravity.google","category":"coding","pricing":"freemium","price_note":"individual $0/mo with weekly agent caps; higher quota via Google AI Pro $19.99/mo or Ultra from $99.99/mo; orgs via Google Cloud consumption","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["gemini-3-1-pro-preview"],"model_routing":"bundled","verdict":"situational","verdict_text":"A real Google coding surface, not Gemini Code Assist: agent models (Gemini, Claude, gpt-oss) inside Antigravity, unlimited tab/command on the free individual plan, paid quota bolted onto Google AI Pro/Ultra. Rate-limit opacity is the tax. Separate from the Gemini consumer app and from Code Assist.","limitations":["Agent quota is weekly/credit-shaped; the public page does not publish a token rate card.","Hitting the cap locks you until refresh unless you buy AI credits or Ultra.","Organization access is Google Cloud / Gemini Enterprise, not a self-serve IDE seat.","Not a replacement for Gemini Code Assist in existing JetBrains/VS Code installs."],"strengths":["Real Google coding surface with agent models from multiple labs.","Free individual tier with unlimited tab and command completions.","Paid quota through familiar Google AI Pro/Ultra subscriptions."],"use_for":["Developers who want a Google-native agentic coding environment.","Individual users who can stay within weekly free-tier agent caps.","Teams already buying Google AI Pro/Ultra for extra quota."],"skip_when":["You need a published token rate card for agent quota.","You cannot risk hitting a weekly cap mid-project.","You want to plug into existing JetBrains/VS Code via Gemini Code Assist."],"receipts":["https://antigravity.google/pricing","https://gemini.google/subscriptions/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"google-cloud-speech-to-text","name":"Google Cloud Speech-to-Text","tagline":"Cloud API that transcribes audio into text in 85+ languages with streaming and speaker diarization.","url":"https://cloud.google.com/speech-to-text","category":"audio","pricing":"freemium","price_note":"V2 API standard recognition starts at $0.016/min; V1 offers 60 free minutes per month, then $0.016–$0.024/min depending on data logging.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Charged per second of successfully processed audio; multiple channels are billed separately."],"receipts":["https://cloud.google.com/speech-to-text","https://cloud.google.com/speech-to-text/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"google-cloud-text-to-speech","name":"Google Cloud Text-to-Speech","tagline":"Cloud API that turns text into natural-sounding speech with hundreds of voices across 75+ languages.","url":"https://cloud.google.com/text-to-speech","category":"audio","pricing":"freemium","price_note":"Free monthly character allowances vary by voice type (e.g., 1M for WaveNet, 4M for Standard); paid per million characters after that.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Pricing is per character and billing starts once monthly free character limits are exceeded."],"receipts":["https://cloud.google.com/text-to-speech","https://cloud.google.com/text-to-speech/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"google-vertex-ai","name":"Google Vertex AI","tagline":"Google Cloud's ML and generative AI platform, now branded Gemini Enterprise Agent Platform.","url":"https://cloud.google.com/vertex-ai","category":"models-api","pricing":"paid","price_note":"Pay-as-you-go for models, compute, storage, and pipelines; new Google Cloud customers get $300 in credits. Agent Platform offers limited free tiers for agent compute, memory, and storage.","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Vertex AI / Agent Platform is a full-stack shop for teams already on Google Cloud, combining Model Garden, tuning, MLOps, and agent tooling. It supports a wide range of Google, third-party, and open models with deep BigQuery and Cloud integrations. The platform is powerful but sprawling; pricing has many moving parts, and the recent rebranding creates navigation friction.","limitations":["Pricing is fragmented across model inference, compute, storage, pipelines, and management fees.","The shift from \"Vertex AI\" to \"Gemini Enterprise Agent Platform\" means docs, SKUs, and URLs are in transition.","Best value is for GCP-centric teams; multi-cloud or simple API-only users may find it heavyweight."],"strengths":["Full-stack Google Cloud ML and generative AI platform.","Supports a wide range of Google, third-party, and open models.","Deep BigQuery and Cloud integrations for GCP-centric teams."],"use_for":["Teams already committed to Google Cloud infrastructure.","Workflows that need model garden, tuning, MLOps, and agent tooling together."],"skip_when":["You are a multi-cloud or simple API-only user who wants a lightweight option.","Fragmented pricing across inference, compute, storage, and fees is a blocker.","The Vertex-to-Agent-Platform rebrand creates navigation friction you cannot afford."],"receipts":["https://cloud.google.com/vertex-ai","https://cloud.google.com/vertex-ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"goose","name":"Goose","tagline":"Open-source, extensible local agent from Block/AAIF — recipes, MCP, BYOK, beyond autocomplete.","url":"https://github.com/aaif-goose/goose","category":"coding","pricing":"free","price_note":"CLI/desktop free (Apache); you pay the model provider; no Goose seat fee","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"A local agent you can actually read: recipes, extensions, MCP, runs on your machine against whatever provider you wire. Quality is the model, not a hosted IDE. Docs URLs have moved with the AAIF donation, so start from the GitHub repo. Not field-run in this pass.","limitations":["No bundled inference; a frontier model bill is still a frontier model bill.","Desktop/docs hosting moved off some Block GitHub Pages paths (404s happen).","You own secrets, tool allowlists, and what the agent can execute locally.","Smaller packaged UX than Cursor/Claude Code; more assembly required."],"strengths":["Open-source local agent with recipes, extensions, and MCP support.","Runs on your machine against whichever provider you wire."],"use_for":["Developers who want a transparent, self-hosted agent and do not mind assembly.","Teams that need BYOK control over models and secrets."],"skip_when":["You want a polished, bundled IDE experience like Cursor or Claude Code.","You expect vendor-hosted inference rather than paying your own model bill.","You cannot manage local secrets, tool allowlists, and execution risk."],"receipts":["https://github.com/aaif-goose/goose","https://goose-docs.ai/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"gpt-image","name":"GPT Image 2","tagline":"OpenAI's image generation/editing model (ChatGPT Images in the app; gpt-image-2 on the API).","url":"https://developers.openai.com/api/docs/models/gpt-image-2","category":"image","pricing":"freemium","price_note":"in ChatGPT: Free/Go limited, Plus expanded, Pro unlimited/faster; API gpt-image-2 $8/1M image-in, $2 cached, $30/1M image-out, $5/1M text-in","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"This is the image job, not a second ChatGPT card. In the app it rides ChatGPT plan caps; on the API it is token-priced (image out is the expensive line). Free API tier does not include it. Quality/size settings change the token count — use OpenAI's calculator, not a flat $/image folklore.","limitations":["API is token-priced; a \"per image\" number depends on quality, size, and edits.","Not available on the OpenAI API Free tier (TPM table: Not supported).","ChatGPT image quotas are plan-gated and slower on Free/Go.","This is not DALL·E 3 and not Sora; video is a different OpenAI SKU."],"strengths":["Native image generation/editing inside the ChatGPT app and API.","Token pricing is published, though the per-image cost depends on settings."],"use_for":["Image generation inside a ChatGPT plan you already use.","API image jobs where token-based billing fits your existing OpenAI workflow."],"skip_when":["You rely on the OpenAI API Free tier — GPT Image 2 is not supported there.","You need a flat per-image rate rather than a token calculator.","Video is the job; that is a different OpenAI SKU."],"receipts":["https://developers.openai.com/api/docs/pricing","https://developers.openai.com/api/docs/models/gpt-image-2","https://chatgpt.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"gpt-sovits","name":"GPT-SoVITS","tagline":"Open-source few-shot voice cloning and text-to-speech toolkit with a built-in WebUI.","url":"https://github.com/RVC-Boss/GPT-SoVITS","category":"audio","pricing":"free","free_tier":true,"open_source":true,"api":false,"self_host":true,"model_routing":"bundled","limitations":["Radar: not field-run in this catalog wave.","Pretrained models are released under a CC-BY-NC license, limiting commercial use without permission."],"receipts":["https://github.com/RVC-Boss/GPT-SoVITS"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"grammarly","name":"Grammarly","tagline":"Writing assistant for grammar, tone and clarity, everywhere you type.","url":"https://www.grammarly.com","category":"writing","pricing":"freemium","price_note":"free tier (100 AI prompts/mo); Pro $12/mo (yearly)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Still the best at catching typos and smoothing tone inside every app. The free tier covers most people. Premium rewrite features are harder to justify now that ChatGPT and Claude do the same job.","limitations":["Suggestions can flatten personal voice into corporate beige.","Premium overlaps with what free LLM chats now do.","Occasional confidently-wrong grammar calls.","Business plan retired; parent company renamed Superhuman (2025-10)."],"strengths":["Best-in-class typo and tone smoothing across apps.","Free tier covers most casual writing needs."],"use_for":["Daily writing where typos and tone slips matter.","Users who want grammar help inside every app, not one chat window."],"skip_when":["You need to preserve a strong personal voice.","Paid rewrite features duplicate free LLM chats you already use.","You cannot tolerate confidently-wrong grammar corrections."],"receipts":["https://www.grammarly.com","https://www.grammarly.com/plans"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"granola","name":"Granola","tagline":"AI notepad for meetings — your rough notes enhanced, no bot in the call.","url":"https://www.granola.ai","category":"productivity","pricing":"freemium","price_note":"free Basic tier; Business $14/user/mo; Enterprise $35/user/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"A notepad that listens and fills in the gaps — no bot in the call, just your rough notes enhanced with transcript detail. Apple-only for now, and the team features are still young, but it solves the meeting-notes problem better than the bot-in-call crowd.","limitations":["Apple-platform-centric; Windows users wait.","System-audio capture has setup quirks on some machines.","Team pricing and admin controls still maturing."],"strengths":["Meeting notes without a bot in the call — your rough notes, filled in.","Solves the notes job better than the bot-in-the-room crowd."],"use_for":["Founders and ICs who already take notes and want the transcript fused in.","Apple-first teams that refuse a meeting bot."],"skip_when":["The team is on Windows.","System-audio capture keeps failing on your machines.","You need mature team admin and pricing."],"receipts":["https://www.granola.ai","https://www.granola.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"grok","name":"Grok","tagline":"xAI's consumer assistant on grok.com — chat, search, Imagine; not the xAI API and not in-X Grok.","url":"https://grok.com","category":"productivity","pricing":"freemium","price_note":"free tier on grok.com; paid SuperGrok upgrades sold in-app (official public $ card is not a stable HTML page — confirm at checkout)","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["grok-4-6"],"verdict":"situational","verdict_text":"grok.com is the standalone chat product. xAI API is a different card; the Grok tab inside X is a different surface. Free exists. Paid SuperGrok is real but x.ai/grok returned 403 here, so this catalog will not invent a dollar figure from blogs. Personality and X-data are the draw; procurement and moderation are the drag.","limitations":["No stable public HTML price table fetched (x.ai/grok 403); treat in-app SuperGrok as authoritative.","Not a substitute for the xAI API rate card (see the xAI API entry).","X-association is a procurement issue for some teams.","Imagine/video features are quota-gated on paid tiers; this card is the chat app, not Grok Imagine as a studio."],"strengths":["Standalone chat product separate from X and the xAI API.","Free tier exists on grok.com.","Personality and X-data integration are the main draw."],"use_for":["Users who want xAI models without building against the API.","People whose workflow already centers on X and wants the same voice."],"skip_when":["You need a stable public price table before purchasing SuperGrok.","xAI API rate card or procurement independence from X is required.","You want Imagine/video as a primary studio rather than a quota-gated chat feature."],"receipts":["https://grok.com","https://docs.x.ai"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"grok-imagine","name":"Grok Imagine","tagline":"xAI's integrated image and video generator inside Grok — up to 2K images and 15-second 1080p videos.","url":"https://x.ai/grok","category":"image","pricing":"freemium","price_note":"Free tier on grok.com; SuperGrok $30/mo unlocks higher Imagine limits and 1080p video; SuperGrok Plus $100/mo for priority and significantly higher usage.","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Grok Imagine is the easiest way for existing Grok users to generate images and short videos without leaving the chat thread. Quality is competitive for quick social assets and memes, but it is tied to Grok's subscription tiers and usage quotas rather than a standalone creative studio. For production pipelines that need API access, batch rendering, or fine-grained control, the xAI API or dedicated media generators are a better fit.","limitations":["No direct API or self-hosted option for Imagine; everything routes through the Grok consumer app.","Video output is limited to 15-second clips and 1080p at the top tier, with quotas that can throttle heavy use.","X/ByteDance-style content moderation and data residency concerns apply for sensitive or enterprise work."],"strengths":["Integrated image and video generation inside the Grok chat thread.","Free tier on grok.com for existing Grok users.","Competitive quality for quick social assets and memes.","Supports up to 2K images and 15-second 1080p videos at the top tier."],"use_for":["Existing Grok users who want quick images and videos without switching apps.","Social assets and memes where speed matters more than pixel-level control."],"skip_when":["You need an API, batch rendering, or fine-grained production control.","Sensitive or enterprise work cannot accept X/ByteDance-style moderation.","You are outside Grok's subscription tiers and quota limits."],"receipts":["https://x.ai/grok","https://x.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"groq","name":"Groq","tagline":"Ultra-fast inference on custom LPU hardware — open models at 500+ tok/s.","url":"https://groq.com","category":"dev-infra","pricing":"freemium","price_note":"free tier; from $0.075/1M tokens (small models)","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"ship","verdict_text":"Sub-second responses really do change what you can build — voice agents, autocomplete, anything real-time. The free tier is big enough to prototype on. Just remember the model catalog is limited to what fits their own hardware.","limitations":["Model catalog limited to what fits their hardware.","Context windows capped below the biggest hosted rivals.","Free-tier throughput limits are tight for production."],"strengths":["Sub-second tokens on their LPUs — voice agents and autocomplete actually feel live.","Free tier is big enough to prototype the latency, not just read a blog post."],"use_for":["Real-time UIs that die at 2-second model latency.","Open-model inference when the model already fits their hardware."],"skip_when":["The model you need is not on their catalog.","You need the largest hosted context windows.","Free-tier throughput would be the production path."],"receipts":["https://groq.com","https://console.groq.com/docs/models","https://console.groq.com/docs/deprecations"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"gumloop","name":"Gumloop","tagline":"AI-native automation platform for building agent-driven workflows visually.","url":"https://www.gumloop.com","category":"automation","pricing":"freemium","price_note":"Pro from $37/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Visual builder for agentic workflows with built-in browser, document, and API nodes. Good for scraping and research pipelines. Still young; expect breaking changes and occasional node limitations.","limitations":["Newer platform with a smaller integration library than Zapier.","Complex workflows can be hard to debug.","Reliability of browser agents varies by site."],"strengths":["Visual builder for agentic workflows with browser, document and API nodes.","Good fit for scraping and research pipelines.","Lower-code entry point than script-heavy automation stacks."],"use_for":["Research and scraping pipelines that need browser and document nodes.","Teams that want a visual automation builder for agentic tasks."],"skip_when":["You need the integration breadth and maturity of Zapier.","Complex branching logic will outgrow the visual editor.","Occasional node limitations and breaking changes would stall production."],"receipts":["https://www.gumloop.com","https://www.gumloop.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"hailuo-ai","name":"Hailuo AI","tagline":"MiniMax consumer video generator ranked high on physics and motion realism, available via API resellers.","url":"https://hailuoai.com","category":"video","pricing":"freemium","price_note":"Consumer plans from ¥105/mo (promo ¥55) to ¥1399/mo, credit-based; API via third-party resellers only; no first-party public price card","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Top-tier physics and motion realism for the price, usually undercutting Runway. The free tier is a solid way to test Hailuo 01/02/H3 before committing to the ¥105–¥1399 credit plans. For developers, the absence of a public first-party video API price means Western access typically routes through resellers.","limitations":["First-party API pricing not public; Western access usually via resellers.","Strict content moderation.","Multiple model tiers (Hailuo 01/02/H3) make the credit system harder to navigate.","Consumer pricing and API pricing are separate, confusing buckets."],"strengths":["Strong physics and motion realism in generated video.","Usually cheaper than Runway for comparable output quality.","Free tier lets you test the Hailuo 01/02/H3 model stack before paying.","Credit-based plans from ¥105/mo up to ¥1399/mo for heavy use."],"use_for":["Short video clips where realistic physics and motion matter most.","Budget-conscious creators comparing against Runway pricing.","Quick experiments on the free tier before buying credits."],"skip_when":["You need a published first-party video API price card.","Strict content moderation would block your subject matter.","You want a single simple credit system across multiple model tiers."],"receipts":["https://hailuoai.com","https://hailuoai.com/pricing","https://www.atlascloud.ai/blog/tips/hailuo-ai-pricing-cost","https://awesomeagents.ai/models/hailuo-02/","https://compareaitools.org/best-ai-video-generator/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"heroui","name":"HeroUI","tagline":"Modern React component library with Tailwind CSS theming.","url":"https://heroui.com","category":"design","pricing":"free","price_note":"free and open source","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"ship","verdict_text":"A reliable component library with good-looking defaults and solid Tailwind integration. Components feel modern and the documentation is easy to follow.","limitations":["Some advanced features require paid templates.","Component count is smaller than mature competitors.","Theming setup takes a moment if you are new to Tailwind v4."],"strengths":["Modern React components with good defaults and solid Tailwind theming.","Docs are easy to follow when you just need a library, not a ritual."],"use_for":["New React apps that want a complete component set with Tailwind.","Teams that prefer a packaged library over copy-paste blocks."],"skip_when":["You need the advanced pieces that only live in paid templates.","The component count is too small versus a mature kit.","Tailwind v4 theming setup is a blocker this week."],"receipts":["https://heroui.com","https://github.com/heroui-inc/heroui"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"heygen","name":"HeyGen","tagline":"AI avatar videos — talking presenters, translation and lip-sync dubbing.","url":"https://www.heygen.com","category":"video","pricing":"freemium","price_note":"free tier (3 videos/mo); Creator $29/mo, Pro from $49/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"The obvious pick for avatar and presenter video — training content, explainers, multilingual dubs. If you need talking heads at scale, nothing else is close. If you don't, it's basically useless.","limitations":["Avatars still read as synthetic on close inspection.","Credit-based pricing (up to 20 credits/min on newer avatars) adds up fast.","Uncanny valley for emotional or high-stakes content."],"strengths":["The obvious pick for avatar and presenter video at scale.","Strong for training content, explainers, and multilingual dubs."],"use_for":["Training videos and explainers that need a talking presenter.","Multilingual dubbing and lip-sync at scale."],"skip_when":["You do not need talking heads; the tool is basically useless without them.","Close inspection will reveal synthetic avatars or uncanny emotional delivery.","Credit-based pricing up to 20 credits/min would blow the budget."],"receipts":["https://www.heygen.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"higgsfield","name":"Higgsfield","tagline":"Multi-model AI video creative suite with camera control, character consistency, and marketing presets.","url":"https://www.higgsfield.ai","category":"video","pricing":"freemium","price_note":"free tier with limited/daily credits; paid from ~$15-19/mo (Starter), $47-59/mo (Plus), $99-129/mo (Ultra)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Higgsfield has evolved from a single video model into a multi-model creative suite (Seedance, Veo, Kling, WAN, Hailuo, etc.) with Cinema Studio camera control, Soul ID character consistency, and Marketing Studio presets. Some models are also available through their native platforms or APIs. Pricing tiers and credit costs shift frequently; verify the live page before budgeting.","limitations":["Now a multi-model platform; quality, speed, and controls depend on the underlying engine.","Premium models can consume credits several times faster than cheaper ones on the same tier.","Standard-tier monthly credits do not roll over.","Heavy single-model users may still get better value from that model's native platform."],"strengths":["Aggregates multiple video models (Seedance, Veo, Kling, WAN, Hailuo) in one suite.","Cinema Studio camera control and Soul ID character consistency for creative direction.","Marketing Studio presets speed up campaign-style video generation."],"use_for":["Creators who want one dashboard instead of juggling separate video-model platforms.","Marketing teams that lean on presets and character-consistent outputs."],"skip_when":["You primarily use one model and can buy credits cheaper on its native platform.","You need predictable pricing that does not shift month to month.","You want credits to roll over instead of expiring each billing period."],"receipts":["https://www.higgsfield.ai","https://higgsfield.ai/blog/best-ai-video-generators-2026"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"honeyhive","name":"HoneyHive","tagline":"Agent observability and evaluation platform for tracing, scoring, and iterating on production agents.","url":"https://www.honeyhive.ai","category":"dev-infra","pricing":"freemium","price_note":"Developer plan is free with 10,000 events per month, up to 5 users, and 30-day retention; Enterprise is custom with self-host and hybrid options.","free_tier":true,"open_source":false,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Developer plan is limited to 10,000 events per month, 5 users, and multi-tenant SaaS hosting only."],"receipts":["https://www.honeyhive.ai","https://docs.honeyhive.ai","https://www.honeyhive.ai/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"horizonx","name":"HorizonX","tagline":"Premium UI kits, Figma templates, and code components for vibecoding.","url":"https://horizonx.so","category":"design","pricing":"paid","price_note":"from $24.99/mo subscription","free_tier":false,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"High-quality assets if you are building slick landing pages and want production-ready Figma and code together. The subscription model only pays off if you ship multiple sites per month.","limitations":["No free tier; everything useful is gated.","Subscription can be expensive for one-off projects.","Heavily focused on marketing sites, not app UI."],"strengths":["Premium Figma templates and code components bundled for landing pages.","Production-ready assets aimed at vibecoding workflows.","Subscription model suits agencies or prolific builders."],"use_for":["Agencies shipping multiple marketing sites per month.","Landing-page builds where matching Figma and code speeds handoff."],"skip_when":["You are building app UI rather than marketing sites.","It is a one-off project and the subscription cost does not amortize.","You need a free tier to evaluate quality first."],"receipts":["https://horizonx.so","https://horizonx.so/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"hugging-face","name":"Hugging Face","tagline":"The GitHub of AI — models, datasets, Spaces and inference endpoints.","url":"https://huggingface.co","category":"dev-infra","pricing":"freemium","price_note":"free; PRO $9/mo; inference pay-as-you-go","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"byok","verdict":"ship","verdict_text":"The default plumbing for open models — weights, datasets, the transformers library, Spaces demos. You probably depend on it already even if you never opened huggingface.co directly. The Hub has hiccups, and sorting good models from noise is still work.","limitations":["Hub reliability incidents affect the whole ecosystem.","Inference Endpoints pricing not always competitive.","Discovering quality models among the noise takes effort."],"strengths":["Default plumbing for open models: weights, datasets, transformers, Spaces.","You probably already depend on it even if you never open the homepage."],"use_for":["Publishing or pulling open weights and datasets.","A first demo Space or a transformers-based prototype."],"skip_when":["Hub incidents would take your whole pipeline down.","Inference Endpoints would be the expensive default.","You need a curated shortlist, not a noisy Hub search."],"receipts":["https://huggingface.co","https://huggingface.co/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"hume-ai","name":"Hume AI","tagline":"Expressive voice / EVI APIs — emotion-conditioned speech, not a generic TTS clone.","url":"https://www.hume.ai","category":"audio","pricing":"freemium","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: pricing and emotion-model claims not reproduced here.","Voice cloning / affect is a consent minefield in production."],"receipts":["https://www.hume.ai","https://dev.hume.ai"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"hunyuan-api","name":"Hunyuan API","tagline":"Tencent Cloud TokenHub API for Hunyuan Hy3, image, video and 3D models (now Tencent Hy).","url":"https://cloud.tencent.com/product/hunyuan","category":"models-api","pricing":"paid","price_note":"Hy3 ~$0.14/$0.56 per 1M tokens (256K context) on TokenHub; Hunyuan 3D from ~1.8 yuan (~$0.27)/generation on TokenHub; older TurboS/T1 and HY2.0 lines retired June 2026","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Hy3 is the current Hunyuan API flagship: cheap, Apache 2.0 open weights, and sold through Tencent TokenHub. TurboS, T1 and the older HY2.0 line were retired in June 2026. Access expanded globally in August 2026 (WorkBuddy, Miora, TokenHub), though real-name verification is still required and some legacy endpoints remain geo-blocked; third-party routers remain common.","limitations":["The hosted API is proprietary; Hy3 weights are released under Apache 2.0 and can be run separately.","New Hunyuan API capabilities are on TokenHub; the old 混元大模型 platform stops new purchases and shuts down 2026-09-30.","TurboS, T1 and the older HY2.0 models were retired in June 2026; do not build new integrations on them.","Access outside China is difficult; real-name verification is required and some endpoints are geo-blocked in the EU/UK/South Korea.","Text, image, video and 3D billing are split across separate TokenHub model endpoints."],"strengths":["Hy3 is cheap on TokenHub with a 256K context window.","Apache 2.0 open weights let you run Hy3 outside the hosted API.","Text, image, video and 3D models are available through one provider."],"use_for":["Cost-sensitive Chinese API access for text and multimodal generation.","Bulk asset generation when you accept ByteDance's trust boundary on BytePlus."],"skip_when":["You need clean USD rate cards or straightforward Western signup.","EU/UK/South Korea geo-blocking is a blocker for your users.","You want to keep building on the retired TurboS/T1/HY2.0 lines."],"receipts":["https://cloud.tencent.com/document/product/1729/97731","https://cloud.tencent.com/developer/article/2674909","https://www.atlascloud.ai/providers/tencent","https://www.apiuspro.cn/tutorial/tencent"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"hunyuan-video","name":"Hunyuan Video","tagline":"Open-weights text/image-to-video generation model from Tencent.","url":"https://github.com/Tencent-Hunyuan/HunyuanVideo","category":"video","pricing":"paid","price_note":"Open weights, self-host infrastructure cost only; HunyuanVideo-1.5 hosted API ~$0.075/sec (480p) via fal.ai; Hunyuan 3D Rapid ~$0.02/image via Atlas Cloud","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"A surprise open-weights release that can run locally and competes with closed video models. The self-hosting cost is real, and the first-party hosted API price was not confirmed at scan time.","limitations":["No native audio; visual-only output.","Self-hosting still requires serious GPU infrastructure.","First-party hosted API is not advertised by Tencent; available through third-party aggregators.","Limited consumer app and support outside China."],"strengths":["Open-weights text/image-to-video model from Tencent.","Can run locally and competes with closed video models.","Available through third-party hosted APIs when self-hosting is too expensive."],"use_for":["Local video generation when you control the hardware and data.","Experiments where open weights matter more than polish."],"skip_when":["You need native audio in the generated video.","You lack the GPU infrastructure for local inference.","You want first-party support and documentation outside China."],"receipts":["https://github.com/Tencent-Hunyuan/HunyuanVideo","https://vuela.ai/models/hunyuan-video/review","https://www.atlascloud.ai/providers/tencent","https://fal.ai/models/fal-ai/hunyuan-video-v1.5/text-to-video"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"hyperbolic","name":"Hyperbolic","tagline":"Open-access AI cloud for on-demand GPUs and pay-as-you-go serverless inference with OpenAI-compatible APIs.","url":"https://hyperbolic.xyz","category":"models-api","pricing":"paid","price_note":"Serverless text inference from $0.10–$4.00 per 1M tokens; dedicated GPU hosting billed hourly by GPU type.","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Basic tier is capped at 60 RPM; higher throughput requires Pro or Enterprise tiers."],"receipts":["https://hyperbolic.xyz","https://www.hyperbolic.ai/docs"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"hyperwrite","name":"HyperWrite","tagline":"AI writing assistant with inline suggestions, research tools, and browser-based agent features.","url":"https://www.hyperwriteai.com","category":"productivity","pricing":"freemium","price_note":"free tier with limited monthly credits on basic AI; Premium $19.99/mo ($16/mo annual); Ultra $44.99/mo ($29/mo annual)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Solid if you want AI writing help inside your existing browser and docs workflow, especially the TypeAhead inline suggestions. Premium overlaps with free ChatGPT/Claude on long-form writing, and the message limits on Premium make it costly for heavy users.","limitations":["Premium is capped at 250 AI messages/month; heavy writers need the pricier Ultra.","The best features (TypeAhead, citations, custom personas) sit behind a paid plan.","Output quality is model-dependent and still needs human fact-checking.","Less compelling than a general-purpose LLM chat if you rarely write in the browser.","Browser extension and agent features are experimental and may break on some sites."],"strengths":["TypeAhead inline suggestions fit browser and docs workflows.","Browser-based AI writing help without leaving your tab.","Solid for short-form copy and inline completion."],"use_for":["Writing in web apps where inline suggestions save time.","Short-form copy that benefits from TypeAhead completion."],"skip_when":["Long-form work exceeds Premium's 250 messages/month.","Free ChatGPT/Claude already covers your writing needs.","You rely on experimental browser agent features."],"receipts":["https://www.hyperwriteai.com","https://www.hyperwriteai.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"hypotenuse-ai","name":"Hypotenuse AI","tagline":"E-commerce and SEO copy generator with bulk article workflows.","url":"https://www.hypotenuse.ai","category":"writing","pricing":"freemium","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: SEO bulk writers are a graveyard-adjacent category.","Output quality not reproduced here."],"receipts":["https://www.hypotenuse.ai"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"ideogram","name":"Ideogram","tagline":"Image generator with the strongest text-in-image rendering.","url":"https://ideogram.ai","category":"image","pricing":"freemium","price_note":"free tier (weekly credits, amount may vary); Plus $15/mo (annual), Pro $42/mo (annual), Team $20/user/mo (annual)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Use this when the image needs readable text — posters, logo drafts, thumbnails, memes. It beats Midjourney and FLUX there. For pure photorealism or art direction, pick one of those instead.","limitations":["Style range narrower than Midjourney.","Free queue can be slow at peak times.","Generations on the free plan are public."],"strengths":["Strongest text-in-image rendering in the category.","Beats Midjourney and FLUX on readable text tasks."],"use_for":["Posters, logo drafts, thumbnails, and memes with readable text.","Any image where text legibility is the hard requirement."],"skip_when":["You need broad style range or photorealism.","Free-queue slowdowns break your workflow.","Public generations on the free plan are a privacy issue."],"receipts":["https://ideogram.ai","https://ideogram.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"inngest","name":"Inngest","tagline":"Durable workflows and event-driven functions for serverless apps.","url":"https://www.inngest.com","category":"dev-infra","pricing":"freemium","price_note":"free Hobby tier; Pro $99/mo","free_tier":true,"open_source":false,"api":true,"self_host":true,"verdict":"ship","verdict_text":"Makes \"sleep for a week then continue\" easy in serverless functions. Great for AI pipelines that need human approval or multi-step state. Self-hosting is not the primary path, so heavy compliance needs require evaluation.","limitations":["The server and CLI use SSPL with delayed Apache-2.0 conversion, not an OSI open-source license today.","Self-hosting story is lighter than Trigger.dev.","Language SDKs are mostly TypeScript/Python.","Event modeling requires upfront design."],"strengths":["Makes sleep-for-a-week-then-continue easy in serverless functions.","Fits AI pipelines that need human approval or multi-step state."],"use_for":["Event-driven TypeScript or Python jobs with durable steps.","Agent pipelines that pause for a human and resume later."],"skip_when":["You need an OSI license today, not SSPL with a delayed conversion.","Self-hosting has to match Trigger.dev's story.","The team is not on TypeScript or Python."],"receipts":["https://github.com/inngest/inngest","https://www.inngest.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"invideo","name":"InVideo","tagline":"AI video generator that turns scripts into publishable clips with stock footage.","url":"https://invideo.io","category":"video","pricing":"freemium","price_note":"Plus from $17/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Fastest way to turn a blog post or script into a faceless YouTube-style video. Stock-heavy output works for explainer content but looks generic, and fine editing control is limited.","limitations":["Generated visuals rely heavily on stock templates.","Voice and avatar quality lag dedicated tools.","Free exports carry a watermark."],"strengths":["Fastest path from a blog post or script to a faceless YouTube-style video.","Stock-heavy output works for explainer content."],"use_for":["Faceless YouTube explainers built from a script or blog post.","Quick social clips where stock footage is acceptable."],"skip_when":["You need fine editing control or unique visuals beyond stock templates.","Voice or avatar quality has to match dedicated tools like HeyGen.","Free exports with a watermark are unacceptable."],"receipts":["https://invideo.io","https://invideo.io/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"inworld-tts","name":"Inworld TTS","tagline":"Realtime text-to-speech API aimed at voice agents and characters, with voice cloning and LLM routing.","url":"https://www.inworld.ai","category":"audio","pricing":"freemium","price_note":"On-Demand free tier includes up to 70 min TTS; paid plans start at $25/mo (Creator), then $100/mo Builder, $300/mo Developer, $1,500/mo Growth, and Enterprise custom.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Inworld TTS-2 is built for low-latency, expressive voice agents and ranks well on TTS benchmarks. The realtime steering and character-friendly voices make it a strong choice for games, companions, and call agents. It is still a narrower ecosystem than ElevenLabs, and the bundled LLM/STT credits can obscure true per-project cost.","limitations":["Pricing is credit-based across TTS, STT, and LLM routing; estimating bills for mixed workloads takes work.","Smaller community and third-party integration library compared to the established audio leaders.","Voice cloning and professional features are gated behind higher tiers or paid add-ons."],"strengths":["Low-latency expressive TTS-2 built for voice agents and characters.","Ranks well on TTS benchmarks for realtime use cases.","On-Demand free tier includes up to 70 minutes of TTS."],"use_for":["Games, companions, and call agents that need expressive realtime speech.","Projects where character-friendly voices matter more than ElevenLabs' ecosystem breadth."],"skip_when":["You need transparent per-project cost across mixed TTS/STT/LLM credits.","Your stack depends on the largest third-party integration library.","Voice cloning is required but the budget only reaches the On-Demand tier."],"receipts":["https://www.inworld.ai","https://www.inworld.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"jasper","name":"Jasper","tagline":"Marketing-focused AI writing platform for teams and brand voice.","url":"https://www.jasper.ai","category":"writing","pricing":"paid","price_note":"$69/seat/mo","free_tier":false,"open_source":false,"api":true,"self_host":false,"verdict":"skip","verdict_text":"The original GPT wrapper, now squeezed between ChatGPT and Claude at $20 and free AI built into every app. Brand-voice features are nice, but not $69/mo nice when the raw models handle most of it.","limitations":["Core value is a wrapper over models you can access directly.","Expensive per seat for what it delivers.","Output needs the same editing as raw ChatGPT output."],"strengths":["Marketing-focused writing templates and brand-voice controls.","Team workflows built around shared content and approvals."],"use_for":["Marketing teams that want guided templates and brand governance.","Organizations that prefer a managed writing platform over raw chat."],"skip_when":["ChatGPT or Claude at lower prices handle the same writing tasks.","The output still needs the same manual editing as raw model output.","$69 per seat per month is hard to justify for a model wrapper."],"receipts":["https://www.jasper.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"redshift","featured":false,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"jenni","name":"Jenni","tagline":"Academic writing assistant for drafting papers, citations, and paraphrasing.","url":"https://jenni.ai","category":"writing","pricing":"freemium","price_note":"Free $0 with daily limits; Plus $12/mo; Pro $29/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Purpose-built for students and researchers who need AI autocomplete, citation management, and PDF chat in one editor. The free tier is heavily capped, and the AI can still hallucinate sources or phrasing. Export and formatting work best on paid plans.","limitations":["Free plan limits daily AI autocomplete, edits, chat, and PDF uploads.","Citation and source suggestions still need manual verification.","Less flexible than a general word processor for non-academic formats."],"strengths":["Purpose-built academic editor with autocomplete, citations, and PDF chat.","One tool for drafting papers, citations, and paraphrasing."],"use_for":["Students and researchers writing papers with citation needs.","PDF chat and paraphrasing inside an academic workflow."],"skip_when":["You need a flexible general word processor.","Free-tier daily limits would stall your writing.","You cannot manually verify AI-generated citations."],"receipts":["https://jenni.ai","https://jenni.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"jules","name":"Jules","tagline":"Google's asynchronous coding agent that clones your repo and works in the background.","url":"https://jules.google.com","category":"coding","pricing":"free","price_note":"public beta is free with usage limits; paid pricing expected after beta","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Jules is the most hands-off Google coding agent: it reads your repo in a cloud VM, plans multi-file changes, and returns a diff. Public beta is currently free with usage limits, but pricing is coming once beta ends. It only works where Gemini is available, and the async workflow means you wait rather than pair live.","limitations":["Only available in regions where Gemini is supported; geo-blocked in some countries.","Asynchronous execution means no real-time pair-programming feedback.","Pricing is undefined beyond the free public beta."],"strengths":["Hands-off Google coding agent that clones a repo and plans multi-file changes.","Public beta is free with usage limits."],"use_for":["Developers in Gemini-supported regions who want async codebase edits.","Quick multi-file diffs when you do not need live pair-programming."],"skip_when":["You are in a region where Gemini is geo-blocked.","You need real-time feedback or synchronous pair-programming.","Undefined post-beta pricing is a blocker for your budget."],"receipts":["https://jules.google.com","https://blog.google/innovation-and-ai/models-and-research/google-labs/jules/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"julius-ai","name":"Julius","tagline":"Chat-based AI workspace for analyzing data, building reports, presentations, and simple web apps.","url":"https://julius.ai","category":"data","pricing":"freemium","price_note":"Free $0/mo; Plus $20/mo ($16/mo billed yearly); Pro $45/mo ($37/mo yearly); Max $200/mo ($166/mo yearly); Business $450/mo ($375/mo yearly); all paid tiers are credit-based","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"A practical no-code analyst for business users who live in spreadsheets and slides. The catch: heavy usage is credit-gated, outputs need human verification, and it is not a substitute for a real data warehouse or BI governance stack.","limitations":["Free tier is limited; meaningful ongoing use requires a paid plan.","Analysis quality depends on the data you upload and the prompts you write.","Credit consumption can be hard to predict for large datasets or frequent reports.","Database connectors and team features are Business-tier only."],"strengths":["Chat-based no-code workspace for spreadsheets, reports, and simple web apps.","Free tier exists; paid tiers start at $16/mo yearly.","Business tier adds database connectors and team features."],"use_for":["Business users who live in spreadsheets and need quick reports or presentations.","Small teams that want a low-code analyst without a BI governance stack."],"skip_when":["You need a real data warehouse or BI governance.","Heavy recurring analysis would blow through unpredictable credits.","Database connectors or team features are required below Business tier."],"receipts":["https://julius.ai","https://julius.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"junie","name":"Junie","tagline":"JetBrains' autonomous coding agent built into IntelliJ, PyCharm, and the JetBrains IDEs.","url":"https://www.jetbrains.com/junie/","category":"coding","pricing":"freemium","price_note":"AI Free limited tier; AI Pro $10/mo, AI Ultimate $30/mo; business seats $20–$60/mo; Enterprise custom","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"byok","verdict":"situational","verdict_text":"Best for teams already living in JetBrains IDEs; it understands project structure and can run, test, and iterate. Outside the JetBrains ecosystem it doesn't exist, and credit-based pricing burns faster than the headline tier suggests.","limitations":["IDE plugin requires a paid JetBrains IDE; CLI works anywhere with BYOK or subscription.","Credit-based pricing burns faster than the tier suggests; heavy use needs AI Ultimate or BYOK CLI.","Can be slow on deeply nested enterprise codebases."],"strengths":["Deep integration with JetBrains project structure, tests and run configs.","Works inside the IDE where Java/Kotlin/Python teams already live.","CLI option extends reach to BYOK or subscription setups."],"use_for":["Teams already committed to JetBrains IDEs for product work.","Codebases where the agent can use existing run and test workflows."],"skip_when":["Your team does not use JetBrains IDEs.","Heavy agent months would push you past the AI Ultimate tier.","Deeply nested enterprise codebases where Junie slows down."],"receipts":["https://www.jetbrains.com/junie/","https://junie.jetbrains.com/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"kilo-code","name":"Kilo Code","tagline":"MIT-licensed coding agent for VS Code, JetBrains, CLI, and cloud — models at provider rates, no markup.","url":"https://kilo.ai","category":"coding","pricing":"freemium","price_note":"individual platform $0; Teams $15/user/mo; Kilo Pass from $19/mo; Gateway at provider rates + 5% credit-purchase fee; cloud compute from $0.33–$1.20/h","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"Open-source agent with a zero-markup gateway is the pitch, and the public price page actually separates platform / inference / cloud compute. Solo with BYOK or local models is free; Teams and Pass are optional. Auto-routing and a 5% credit fee still need watching. Not a Ship without a repo run.","limitations":["Three meters (platform, inference, cloud compute) make a real bill three line items.","Credit top-ups carry a 5% processing fee on the published card.","Cloud agents bill per second on top of tokens; idle VMs are not free.","\"Free models\" via Auto Free depend on whatever the gateway currently offers."],"strengths":["Open-source agent with a zero-markup gateway at provider rates.","Public pricing separates platform, inference and cloud compute meters.","Solo use with BYOK or local models is free."],"use_for":["Solo developers who want a free coding agent with their own API keys.","Teams that prefer an open, self-hostable agent over a black-box editor."],"skip_when":["Three separate meters make your bill unpredictable.","A 5% credit top-up fee and Auto-routing uncertainty bother you.","You are not willing to run and verify the open-source repo yourself."],"receipts":["https://kilo.ai/pricing","https://kilo.ai/docs/gateway/usage-and-billing","https://github.com/Kilo-Org/kilocode"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"kimi-api","name":"Kimi API","tagline":"Moonshot OpenAI-compatible API — K3 1M context, K2.7 Code, pay-as-you-go after a $1 top-up.","url":"https://platform.kimi.ai","category":"models-api","pricing":"paid","price_note":"no free API calls; K3 unlocked after ≥$1 top-up; K3 $0.30 cache-hit / $3 input / $15 output per 1M; K2.7 Code $0.19 / $0.95 / $4 per 1M; K2.6 $0.16 / $0.95 / $4","free_tier":false,"open_source":false,"api":true,"self_host":false,"models_used":["kimi-k3","kimi-k2-7-code"],"model_routing":"locked","verdict":"situational","verdict_text":"The Moonshot token meter, not Kimi Code CLI and not kimi.com chat. OpenAI SDK against api.moonshot.ai. K3 is flat $3/$15 with a cheap cache hit; coding SKU is K2.7 Code. Rate limits follow cumulative top-up. China-hosted — same compliance question as other CN labs. Not field-called here.","limitations":["K3 is gated behind a successful top-up (docs: minimum $1); this is not a free playground.","Rate limits scale with lifetime recharge, not with a published unlimited Pro seat.","Docs currently warn that web_search is being updated — don't build on that tool yet.","Separate from Kimi Code membership (~$19–$39) and from the MIT CLI."],"strengths":["OpenAI-compatible API with K3 1M context and K2.7 Code SKU.","Cheap cache-hit pricing on K3.","Coding SKU is clearly separated from general chat."],"use_for":["Existing OpenAI SDK projects that want to route to Moonshot models.","High-context tasks that fit K3's 1M token window."],"skip_when":["You need a free playground before committing budget.","China-hosted compliance is a hard no for your data.","You want to build on the web_search tool before it stabilizes."],"receipts":["https://platform.kimi.ai/docs/pricing/chat-k3","https://platform.kimi.ai/docs/overview"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"kimi-code","name":"Kimi Code","tagline":"Open-source terminal coding agent from Moonshot AI with Anthropic-compatible gateway and MCP support.","url":"https://www.kimi.com/code","category":"coding","pricing":"freemium","price_note":"CLI is free/MIT; hosted coding membership ~$19–$39/mo depending on region; API billed separately at Kimi API rates (K3 $3/$15 per 1M)","free_tier":true,"open_source":true,"api":true,"self_host":true,"models_used":["kimi-k3"],"model_routing":"byok","verdict":"ship","verdict_text":"A credible open-source Claude Code alternative. The CLI is free and the Anthropic-compatible gateway means it drops into existing toolchains. Costs can climb if you lean on hosted inference, so BYOK is the smart setup.","limitations":["Pricing story is complicated; \"$19 flat\" is misleading for heavy users.","Newer and less polished ecosystem than Claude Code.","Data goes to Moonshot unless self-hosted/BYOK.","Multiple regional plans make comparison confusing."],"strengths":["Credible open-source Claude Code alternative with an Anthropic-compatible gateway.","CLI is MIT; drops into existing toolchains without a new IDE.","BYOK is the honest cheap path — hosted membership is optional."],"use_for":["Terminal agents at Chinese-lab prices when you will bring your own keys.","Teams that already speak Anthropic APIs and want a Moonshot route."],"skip_when":["You believed \"$19 flat\" covers heavy hosted inference — it does not.","You need Claude Code's polish and ecosystem more than the price cut.","Code cannot go to Moonshot and you will not self-host."],"receipts":["https://www.kimi.com/code","https://github.com/MoonshotAI/kimi-code","https://moonshotai.github.io/kimi-code/en/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"kiro","name":"Kiro","tagline":"AWS spec-driven IDE — requirements and tasks before code, billed in monthly credits.","url":"https://kiro.dev","category":"coding","pricing":"freemium","price_note":"Free $0 (50 credits); Pro $20/user/mo (1,000 credits); Pro+ $40; Pro Max $100; Power $200; overage $0.04/credit","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Amazon's IDE bet after Q: you approve a spec, then the agent implements. Credit math is public ($20 → 1k credits, $0.04 overage). Unused credits do not roll over on the FAQ. Not Amazon Q, not Bedrock console. Untested here on a real spec cycle.","limitations":["Monthly credits do not roll over (stated on the pricing FAQ).","Free-tier model mix (Sonnet 4.5 + open weights) is narrower than paid Auto/Opus.","Premium models are geo-restricted; not every country gets the full picker.","Spec-first workflow is overhead if you just wanted tab-complete."],"strengths":["Spec-first workflow: approve requirements before the agent writes code.","Public credit math ($20 = 1,000 credits, $0.04 overage).","AWS-aligned IDE with deep cloud integration potential."],"use_for":["Teams that want explicit spec approval before agentic implementation.","AWS shops already using Amazon Q/Bedrock adjacent tooling.","Developers who prefer structured planning over tab-complete."],"skip_when":["You just want fast tab completion without spec overhead.","Geo-restrictions block the premium models you need.","Non-rolling monthly credits would waste your budget."],"receipts":["https://kiro.dev/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"kling","name":"Kling AI","tagline":"Kuaishou's video model — strong motion and physics at aggressive prices.","url":"https://klingai.com","category":"video","pricing":"freemium","price_note":"free daily credits; paid from ~$10/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"The best price-performance pick in AI video right now. Motion and prompt adherence rival Runway for a fraction of the cost, and free daily credits keep experimentation cheap. It's a Chinese service, so expect strict moderation, occasional queues, and possible compliance pushback.","limitations":["Content moderation is strict and occasionally arbitrary.","Generation queues stretch on lower tiers.","Data policies of a Chinese vendor may fail some compliance reviews."],"strengths":["Best price-performance in AI video right now: motion that rivals Runway for less.","Free daily credits keep experimentation cheap."],"use_for":["Short clips where motion and prompt adherence matter more than a US vendor.","Iteration on video ideas without Runway credit math."],"skip_when":["A Chinese vendor's data policy fails the compliance review.","Strict or arbitrary moderation would kill the brief.","Lower-tier queues would miss the deadline."],"receipts":["https://klingai.com","https://uragent.org/tools/kling_ai/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"kokoro-tts","name":"Kokoro","tagline":"Small open TTS model (82M) that runs locally at usable quality for the size.","url":"https://github.com/hexgrad/kokoro","category":"audio","pricing":"free","free_tier":true,"open_source":true,"api":false,"self_host":true,"limitations":["Radar: we have not bench'd Kokoro against ElevenLabs on a studio job.","Local GPU/CPU story depends on the runtime you wrap it in."],"receipts":["https://github.com/hexgrad/kokoro"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"krea","name":"Krea","tagline":"Real-time AI creative suite for images, video, and 3D.","url":"https://krea.ai","category":"image","pricing":"freemium","price_note":"free 100 units/day; Basic $9/mo, Pro $35/mo, Max $105/mo (yearly ~$5/$21/$63); Business flat up to 50 seats, Enterprise custom","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"The live canvas and style controls make it the fastest way to iterate on a look. Great for concept art, logos, and product shots. Print-ready resolution and exact brand fidelity still need work.","limitations":["Highest resolutions and commercial usage require paid plans.","Style consistency across sessions can drift.","Less suited for precise text-in-image tasks."],"strengths":["Live canvas and style controls make iterating a look faster than batch generators.","Strong on concept art, logos and product-shot exploration."],"use_for":["Concepting a visual direction in real time.","Product and brand sketches before a print-ready pass elsewhere."],"skip_when":["You need print resolution or commercial usage on the free plan.","Exact brand fidelity has to survive across sessions.","The job is precise text-in-image."],"receipts":["https://krea.ai","https://krea.ai/pricing","https://pikes.ai/blog/krea-ai-review-2026"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"krisp","name":"Krisp","tagline":"AI noise cancellation, accent conversion, and meeting notes for voice calls.","url":"https://krisp.ai","category":"audio","pricing":"freemium","price_note":"7-day free trial; paid Meeting AI plans; Call Center AI and Developer APIs are custom-priced","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Still the benchmark for real-time noise cancellation, now expanded with AI meeting notes and accent conversion. The free trial is short; paid plans add transcription, CRM sync, and admin controls. Best for remote workers and call centers, not casual users.","limitations":["Full feature set requires a paid plan after a 7-day trial.","Meeting assistant and call-center products are priced separately.","Accent conversion and some transcription features are server-side, raising privacy considerations."],"strengths":["Real-time noise cancellation that remains the category benchmark.","Paid plans add meeting notes, accent conversion, and admin controls.","Strong fit for remote workers and call-center environments."],"use_for":["Remote workers who need clean audio on calls in noisy environments.","Call centers that want accent conversion and CRM-linked meeting notes."],"skip_when":["You are a casual user who only needs occasional noise suppression.","Server-side accent conversion raises privacy concerns for your calls.","The 7-day free trial is too short to evaluate the full paid feature set."],"receipts":["https://krisp.ai","https://krisp.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"lambda","name":"Lambda","tagline":"Cloud GPU instances and API inference for training and deploying AI models.","url":"https://lambda.ai","category":"models-api","pricing":"paid","price_note":"pay-as-you-go GPU and API pricing","free_tier":false,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Reliable GPU cloud with competitive pricing for training and inference. Good for teams that need dedicated hardware without hyperscaler complexity. The API product is newer and less mature than compute instances.","limitations":["API service is newer than the GPU cloud.","Support and tooling lag AWS/GCP.","Pricing for reserved capacity requires planning."],"strengths":["Reliable GPU cloud with competitive pricing for training and inference.","Dedicated hardware without hyperscaler complexity and lock-in."],"use_for":["Teams that need dedicated GPU instances for training or inference.","Workloads where predictable hardware beats serverless token pricing."],"skip_when":["You need a mature managed API rather than raw compute instances.","Your team lacks the capacity to plan reserved capacity and ops.","AWS/GCP-level support and tooling are non-negotiable."],"receipts":["https://lambda.ai","https://lambda.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"lancedb","name":"LanceDB","tagline":"AI-native multimodal lakehouse unifying data curation, retrieval, and training.","url":"https://www.lancedb.com","category":"data","pricing":"freemium","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Cloud and enterprise pricing is not published on the site; sales contact is required."],"receipts":["https://www.lancedb.com"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"landingfolio","name":"Landingfolio","tagline":"Directory of landing page designs, templates, and components.","url":"https://landingfolio.com","category":"design","pricing":"freemium","price_note":"free directory; paid templates","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Use it as a mood board before you design a landing page — the free directory is solid, and the paid templates are a faster start when you want a finished layout. Don't expect copy-paste components or consistent quality across every template.","limitations":["Free tier is browse-only.","Template quality is mixed.","Not a component library for copy-paste."],"strengths":["Free directory of landing-page patterns and inspiration.","Paid templates can speed up a design start."],"use_for":["Browsing landing-page patterns before committing to a design.","Teams that want a template head start rather than a blank canvas."],"skip_when":["You need a copy-paste component library.","Template quality consistency is critical.","The free browse-only tier does not give you usable assets."],"receipts":["https://landingfolio.com","https://landingfolio.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"langflow","name":"Langflow","tagline":"Open visual framework for RAG and multi-agent graphs, now under IBM/DataStax.","url":"https://www.langflow.org","category":"automation","pricing":"freemium","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: no field run. IBM/DataStax gravity may change the OSS cadence.","Graphs screenshot well and fail in production the same way all visual LLM builders do."],"receipts":["https://www.langflow.org","https://github.com/langflow-ai/langflow"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"langfuse","name":"Langfuse","tagline":"Open-source LLM observability — traces, prompts, evals; MIT self-host or Cloud from Hobby.","url":"https://langfuse.com","category":"dev-infra","pricing":"freemium","price_note":"self-host OSS free (MIT); Cloud Hobby $0 (50k units/mo, 2 users); Core $29/mo; Pro $199/mo; Enterprise from $2,499/mo; overage from $8/100k units","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"langfuse.com is live (the 404 rumor is stale). Cloud is unit-metered (traces + observations + scores); Hobby is a real free POC; Core at $29 is the production on-ramp. Self-host OSS is full core features, Enterprise is the RBAC/audit layer plus ClickHouse commercial. Not a field install here.","limitations":["A \"unit\" is every trace, span, generation, and score — verbose agents get expensive.","Hobby hard-caps at 50k units and 2 users; no overage.","SSO/RBAC on Cloud is Pro+$300 Teams add-on or Enterprise, not Core.","Self-host Enterprise pricing is additive to ClickHouse Cloud/BYOC."],"strengths":["MIT-licensed self-host option with full core observability features.","Cloud Hobby is a real free POC for small teams.","Covers traces, prompts, evals, and unit-metered usage in one tool."],"use_for":["Teams that need LLM observability and can self-host for compliance.","Small POCs that fit inside Hobby's 50k units and 2 users."],"skip_when":["Verbose agents would blow through the per-unit meter quickly.","You need SSO/RBAC without paying Pro, Teams add-on, or Enterprise.","You expect overage on the free Hobby tier."],"receipts":["https://langfuse.com/pricing","https://langfuse.com/pricing-self-host","https://github.com/langfuse/langfuse"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"langgraph","name":"LangGraph","tagline":"Graph-based agent orchestration from LangChain — stateful and controllable.","url":"https://www.langchain.com/langgraph","category":"agents","pricing":"freemium","price_note":"framework free open-source; LangSmith Plus $39/seat/mo","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"ship","verdict_text":"Production agents need explicit state, human checkpoints, and replay — LangGraph delivers all three. The learning curve is real, mostly because LangChain leaks through the abstractions. Overkill for a simple single-agent script; hard to beat once a workflow has to stay observable.","limitations":["Learning curve is real; LangChain abstractions leak through.","Docs sprawl across the whole LangChain ecosystem.","Overkill for simple single-agent tasks.","LangGraph Platform renamed LangSmith Deployment; usage billed in LCU/LSU."],"strengths":["Explicit state, human checkpoints and replay — what production agents actually need.","Hard to beat once a workflow has to stay observable."],"use_for":["Multi-step agents that must pause, resume and be inspected.","Teams already in the LangChain orbit who need a real graph, not a chain demo."],"skip_when":["The job is a simple single-agent script.","You will not climb the LangChain leak-through learning curve.","Docs sprawl across the ecosystem is a non-starter."],"receipts":["https://github.com/langchain-ai/langgraph"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"languagetool","name":"LanguageTool","tagline":"Open-source multilingual grammar, style, and spelling checker.","url":"https://languagetool.org","category":"writing","pricing":"freemium","price_note":"Free browser/editor with basic checks; Premium unlocks advanced style, grammar, and higher limits","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"The strongest open-source option for multilingual grammar and style checking, with a solid free browser add-on. Premium unlocks deeper style rules and higher limits, but the core engine is self-hostable. Best for privacy-conscious users and teams that want EU-based tooling.","limitations":["Free tier has a 2,000-character limit per check and fewer style rules.","Premium paraphrasing is capped even on paid plans.","Self-hosted setup requires technical effort and your own infrastructure."],"strengths":["Strongest open-source option for multilingual grammar and style checking.","Solid free browser add-on for quick checks.","Core engine is self-hostable for privacy-conscious teams."],"use_for":["Multilingual writers who need an open-source grammar/style checker.","Privacy-conscious users and teams wanting EU-based tooling.","Teams with the ops capacity to self-host the core engine."],"skip_when":["You need unlimited style checks without the Premium cap.","The free 2,000-character limit per check breaks your workflow.","You cannot invest the technical effort to self-host."],"receipts":["https://languagetool.org","https://languagetool.org/premium"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"langwatch","name":"LangWatch","tagline":"Open-source LLMOps platform for agent observability, evaluations, simulations, and prompt management.","url":"https://langwatch.ai","category":"dev-infra","pricing":"freemium","price_note":"Developer plan is free forever with 50,000 events per month, 14-day data access, and 2 users; Growth is $34 per core-seat per month; Enterprise is custom.","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Developer plan limits data access to 14 days, 2 users, and 3 scenarios, 3 simulations, and 3 custom evals each."],"receipts":["https://langwatch.ai","https://docs.langwatch.ai","https://langwatch.ai/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"leonardo-ai","name":"Leonardo AI","tagline":"Production-oriented image suite with fine-tuning, canvas editing and team features.","url":"https://leonardo.ai","category":"image","pricing":"freemium","price_note":"free tier (150 tokens/day); Essential $12/mo, Premium $30/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"A practical all-in-one for game assets and marketing teams that want fine-tuned styles without managing GPUs. Quality is decent, not class- leading; you are mostly paying for the workflow and canvas tools.","limitations":["Raw model quality a notch below Midjourney and FLUX.","Token math across features is opaque.","Acquired by Canva; roadmap now tied to that orbit."],"strengths":["All-in-one image suite with fine-tuning, canvas editing, and team features.","Practical for game assets and marketing teams that want style consistency without GPU ops.","Workflow and canvas tools are the main value over raw model quality."],"use_for":["Game asset pipelines that need quick style iteration and team sharing.","Marketing teams generating campaign visuals without managing model infrastructure."],"skip_when":["Raw image quality must match Midjourney or FLUX.","Opaque token math across features would frustrate budgeting.","You are worried about roadmap changes after the Canva acquisition."],"receipts":["https://leonardo.ai","https://leonardo.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"lex","name":"Lex","tagline":"Minimal AI-native writing app — Google Docs with a co-writer built in.","url":"https://lex.page","category":"writing","pricing":"freemium","price_note":"free tier; Pro from ~$16/mo (sources differ $16-18/mo, $96-145/yr — confirm on the Stripe pricing table)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"A clean, focused long-form writing app. Inline prompts and version history help if drafting is your main job. Overkill for occasional paragraph help, and ChatGPT Canvas covers part of it free.","limitations":["A subscription on top of what ChatGPT Canvas partially does free.","Far smaller feature set than Notion or Google Docs.","Mobile experience is thin."],"strengths":["Minimal, focused writing surface for long-form drafting.","Inline prompts and version history are built around writing workflows.","Cheaper entry point than sprawling all-in-one docs tools."],"use_for":["Writers and editors whose main job is drafting long articles or newsletters.","Teams that want a Google Docs-like co-writer without Notion's complexity."],"skip_when":["You only need occasional paragraph rewrites that ChatGPT Canvas already handles.","You need a capable mobile writing experience.","You rely on databases, wikis, or advanced document features."],"receipts":["https://lex.page","https://lex.page/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"lindy","name":"Lindy","tagline":"AI agent platform that builds automations from natural language.","url":"https://www.lindy.ai","category":"automation","pricing":"paid","price_note":"7-day free trial; Plus $29.99/user/mo, Pro $99.99, Max $199.99 (credits pooled per workspace)","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Turns plain-English descriptions into agent workflows across email, calendar, and apps. Useful for non-technical operators, but the magic fades when edge cases need explicit handling.","limitations":["Limited control over precise step-by-step logic.","Pricing escalates with active agents and tasks.","Integrations are narrower than Zapier/Make."],"strengths":["Builds automations from plain-English descriptions.","Accessible to non-technical operators for email, calendar and app workflows."],"use_for":["Non-technical teams automating routine email and calendar tasks.","Quick workflows where edge cases are rare and tolerance for approximation is high."],"skip_when":["You need precise step-by-step control over the automation logic.","Active agents and tasks would push pricing up quickly.","You need the integration breadth of Zapier or Make."],"receipts":["https://www.lindy.ai","https://www.lindy.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"linear","name":"Linear","tagline":"Issue tracker with a first-party agent platform — issues, cycles, and an agent that can open coding sessions.","url":"https://linear.app","category":"productivity","pricing":"freemium","price_note":"Free $0 (2 teams, 250 issues); Basic $10/user/mo billed yearly; Business $16/user/mo billed yearly; coding sessions spend AI credits","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"The tracker teams actually keep using. Agent platform and Linear Agent ship on Free; coding sessions are starred as AI-credit extras. Basic/Business are yearly-priced on the public page. This is project ops, not an IDE. No production workspace run in this pass.","limitations":["Free caps at 250 issues and 2 teams.","Coding sessions and some agent features consume AI credits on top of the seat.","Business/Enterprise features (Asks, Insights, SAML) are not on Basic.","Not a substitute for GitHub Issues if the team already lives in PRs only."],"strengths":["Issue tracker teams actually keep using.","First-party agent platform and Linear Agent ship on the free plan.","Coding sessions connect tickets to actual coding work."],"use_for":["Teams that need issue tracking plus a built-in coding agent.","Workflows that combine cycles, issues, and coding sessions."],"skip_when":["You need more than 250 issues or 2 teams on the free plan.","You want coding sessions without buying AI credits on top of the seat.","Your team already lives entirely inside GitHub Issues/PRs."],"receipts":["https://linear.app/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"linear-mcp","name":"Linear MCP","tagline":"Linear's hosted MCP — issues, projects, comments. Read-write at /mcp, read-only at /readonly.","url":"https://linear.app/docs/mcp","category":"mcp","pricing":"free","price_note":"included with a Linear account; hosted https://mcp.linear.app/mcp; readonly https://mcp.linear.app/mcp/readonly; OAuth 2.1 or API key","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"Linear's own remote MCP: Streamable HTTP, OAuth 2.1, find/create/update on issues, projects, comments. A /readonly URL exists so you do not have to hand the agent a write token. No extra MCP seat — you already pay Linear.","limitations":["Hosted-only; your Linear workspace leaves the machine.","Default /mcp is read-write. Use /mcp/readonly if the agent should not file issues.","No separate MCP price, but Linear Free is still 2 teams / 250 issues."],"strengths":["Official hosted server at mcp.linear.app, not a community wrapper.","Separate /readonly URL; read OAuth scope cannot reach write APIs.","Native in Claude, Cursor, Codex, VS Code, Windsurf, Zed; mcp-remote for older clients."],"use_for":["Coding agents that triage Linear issues or open tickets from a session.","Hosts that speak remote MCP and should not run a local Linear binary."],"skip_when":["The team lives in GitHub Issues — use GitHub MCP, not this.","You will not give an agent write on the tracker — then use /mcp/readonly only.","You needed self-hosted Linear; this server is Linear's cloud."],"receipts":["https://linear.app/docs/mcp","https://mcp.linear.app/mcp"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":false},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"litellm","name":"LiteLLM","tagline":"Open-source OpenAI-compatible gateway — 100+ providers, spend keys, optional Enterprise license.","url":"https://www.litellm.ai","category":"dev-infra","pricing":"freemium","price_note":"OSS proxy/SDK free (self-host); SSO free up to 5 users; Enterprise license is usage-quoted (30-day trial, no public $ table); 24/7 SLA extra","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"The self-host cousin of OpenRouter: one OpenAI-shaped interface, virtual keys, budgets, fallbacks. OSS already does a lot. Enterprise is SSO/SCIM, audit, multi-region, and a license key — docs say pricing is usage-based and sales-quoted, so this card will not copy third-party $250/$30k blogs. You still pay every upstream model bill.","limitations":["docs.litellm.ai/docs/pricing 404s; Enterprise dollars are not on a public SKU page.","SSO beyond 5 users requires the commercial license.","You run Postgres/Redis/the proxy; the \"free\" gateway is not zero-ops.","Several built-in guardrail callbacks are Enterprise-gated, not OSS."],"strengths":["One OpenAI-shaped interface to more than a hundred providers.","Virtual keys, budgets and fallbacks in the OSS gateway.","Self-hostable proxy and SDK without a seat license."],"use_for":["Teams routing multiple model providers through one API shape.","Self-hosted deployments that need spend controls and virtual keys.","Evaluation setups that want to swap providers without code changes."],"skip_when":["You need a public, fixed Enterprise price list.","You want a zero-ops managed service without running Postgres/Redis.","SSO beyond five users is required without the commercial license."],"receipts":["https://docs.litellm.ai/docs/enterprise","https://github.com/BerriAI/litellm"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"llamaindex","name":"LlamaIndex","tagline":"Data framework and document parsing platform for building context-augmented generative AI apps.","url":"https://www.llamaindex.ai","category":"data","pricing":"freemium","price_note":"LlamaParse free plan includes 10,000 credits per month (~1000 pages).","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Cloud APIs such as LlamaParse are metered beyond the free credit allowance."],"receipts":["https://www.llamaindex.ai"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"llamaparse","name":"LlamaParse","tagline":"LlamaIndex hosted document parser — credits per page, not an OSS engine.","url":"https://www.llamaindex.ai","category":"data","pricing":"freemium","price_note":"10K credits/mo on Free (hard stop); 1,000 credits = $1.25 (NA/EU); v2 Fast 1 / Cost-effective 3 / Agentic 10 / Agentic Plus 45 credits/page; Starter 40K incl. (PAYG to $500/mo); Pro 400K (PAYG to $5,000/mo); seat $ not in this HTML extract","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"The PDF hammer in LlamaIndex tutorials is a commercial API. Official FAQ: LlamaParse is not open source; the GitHub repo is a client. Free is 10K credits/month then 402. $1.25 per 1,000 credits makes Fast ~$1.25/1k pages and Agentic Plus much more. VPC exists for enterprises, not a docker OSS story. Confirm Starter/Pro seat dollars in the console — they were not in this HTML extract.","limitations":["Official FAQ: LlamaParse itself is not OSS; don't confuse the Python client with the parser.","Free plan has no PAYG — extra usage returns 402 until you upgrade.","Agentic Plus is 45 credits/page (~$0.056) before extract add-ons.","Seat prices for Starter/Pro were not rendered as dollars in the public HTML; budget from credits + PAYG caps."],"strengths":["Hosted PDF parser tightly integrated with the LlamaIndex ecosystem.","Per-page credit pricing is transparent enough to model costs.","Enterprise VPC option exists for stricter data boundaries."],"use_for":["RAG pipelines parsing PDFs and documents inside LlamaIndex workflows.","Teams that prefer a managed parser over self-hosting an OSS alternative."],"skip_when":["You need an open-source parser you can self-host for free.","Free 10K credits/month would cap before your volume finishes.","Budget requires confirmed seat dollars not shown in public pricing."],"receipts":["https://developers.llamaindex.ai/llamaparse/general/pricing/","https://www.llamaindex.ai/pricing","https://github.com/run-llama/llama_parse"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"lovable","name":"Lovable","tagline":"Prompt-to-full-stack app builder — credits for builds, GitHub/GitLab sync, hosted backends.","url":"https://lovable.dev","category":"coding","pricing":"freemium","price_note":"Free $0 (5 daily build credits, 30/mo cap); Pro from $25/mo (100 credits); Business from $50/mo; annual Pro $250/yr","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"The polished sibling of Bolt: credits, roles, custom domains, Git sync. Pro is $25 for 100 monthly credits plus 5 daily builds; Business is the same credit floor with SSO. Daily credits die at midnight UTC. Cloud and AI grants are extra meters. No production app shipped from here.","limitations":["Daily build credits reset 00:00 UTC and do not roll over.","Free cannot download code or hide the Lovable badge.","Business is governance (SSO, security center), not extra base credits.","Credit spend is opaque per prompt — a messy refactor can torch the month."],"strengths":["Polished prompt-to-full-stack builder with GitHub/GitLab sync and custom domains.","Pro tier adds 100 monthly credits on top of daily free builds.","Credit-based pricing separates build consumption from seat count."],"use_for":["Rapid MVPs and internal tools that need Git sync and a hosted backend.","Teams that want a polished Bolt alternative inside the React/full-stack ecosystem."],"skip_when":["You need free code download or a white-label build without the Lovable badge.","Daily credit resets at midnight UTC would break your workflow.","You cannot tolerate opaque credit burn per prompt."],"receipts":["https://docs.lovable.dev/introduction/subscription-plans","https://lovable.dev/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"ltx-studio","name":"LTX Studio","tagline":"Lightricks' AI film studio — storyboards, LTX-2.5 video, credit-metered plans from a free pile.","url":"https://ltx.studio","category":"video","pricing":"freemium","price_note":"Free $0 with 800 one-time credits (personal use only); Lite $15/mo (8k credits); Standard $35/mo (28k); Pro $125/mo (110k); yearly ~20% off ($12 / $28 / $100)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"A timeline/storyboard product, not a single-clip model playground. Free is 800 credits once and personal-use licensed. Commercial use starts at Standard. Veo and Kling sit on higher SKUs; LTX's own 2.5 is on all plans. Credit spend per shot is the real bill. Not field-cut here.","limitations":["Free credits are one-time, not monthly, and cannot be used for paid/business work.","Third-party models (Veo, Kling, Seedance, GPT Image) are gated to Standard/Pro, not Lite.","Promo/Cyber Monday strikethroughs on the page — budget the non-promo monthly/yearly numbers.","Collaboration beyond three people per project is Enterprise."],"strengths":["Full film-studio timeline and storyboard workflow, not a single-clip model.","LTX-2.5 video is available on every plan.","Clear credit-metered plans with ~20% annual discount."],"use_for":["Video projects that need storyboards, shots, and a timeline.","Creators ready to start at Standard for commercial use.","Teams that want access to Veo/Kling on higher SKUs."],"skip_when":["You need free credits for commercial or ongoing monthly work.","You want third-party models on the Lite plan.","Collaboration beyond three people per project is required without Enterprise."],"receipts":["https://ltx.studio/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"lucide","name":"Lucide","tagline":"Open-source icon library with clean, consistent SVG icons.","url":"https://lucide.dev","category":"design","pricing":"free","price_note":"ISC license","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"ship","verdict_text":"The default icon set for modern React and web projects. The icons are consistent, the packages are well maintained, and it integrates with shadcn/ui out of the box.","limitations":["Style is uniform; less useful if you need highly custom iconography.","Some niche categories are thinner than larger sets.","Icon naming can differ from other libraries."],"strengths":["Default icon set for modern React: consistent SVGs, well-maintained packages.","Drops into shadcn/ui without a second icon system."],"use_for":["App chrome and marketing UI that should look like one family.","React projects already on shadcn or a similar Tailwind stack."],"skip_when":["You need highly custom iconography, not a uniform stroke set.","A niche category is missing versus a larger library.","Icon names have to match another set you already standardized on."],"receipts":["https://lucide.dev","https://github.com/lucide-icons/lucide"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"luma-dream-machine","name":"Luma Dream Machine","tagline":"Luma's fast ideation video model with start/end keyframe control — inside its creative-agents platform.","url":"https://lumalabs.ai","category":"video","pricing":"freemium","price_note":"free tier (watermarked); from $30/mo (Luma Agents Plus)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Use it for fast storyboard iteration. Keyframe control and quick generation make it a useful sketchpad, not a final renderer. Realism and motion sit below Veo, Runway and Kling, and the paid lineup still straddles legacy Dream Machine and Luma Agents since 2026-03.","limitations":["Realism and fine motion below Veo and Runway.","Limited clip duration per generation.","Far fewer post-generation editing tools than Runway.","Two plan ladders coexist (legacy Dream Machine and Luma Agents since 2026-03); clean export from paid only."],"strengths":["Fast generation and start/end keyframe control for storyboard iteration.","Free tier gives watermarked previews without upfront spend.","Quick sketchpad for visual ideas before committing to a final renderer."],"use_for":["Fast storyboard iteration where speed matters more than final realism.","Creatives who want keyframe-controlled video prototypes."],"skip_when":["You need final-render realism and fine motion matching Veo or Runway.","Post-generation editing tools are a core part of your workflow.","The coexistence of legacy Dream Machine and Luma Agents plans is confusing."],"receipts":["https://lumalabs.ai","https://lumalabs.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"magic-ui","name":"Magic UI","tagline":"Animated React components and effects for design engineers.","url":"https://magicui.design","category":"design","pricing":"free","price_note":"MIT open source","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"ship","verdict_text":"Excellent for landing-page sparkle without drowning in animation code. The components copy-paste cleanly into shadcn/ui projects and feel consistent.","limitations":["Effects can be overused; restraint is required.","Some components assume Tailwind and Framer Motion are already in place.","Not a full app component library."],"strengths":["Landing-page sparkle without drowning in animation code.","Copy-paste blocks that sit cleanly on shadcn/ui."],"use_for":["Marketing pages that need a few strong effects, not a full app kit.","Design-engineer workflows already on Tailwind and shadcn."],"skip_when":["The page would overuse effects without an editor to restrain them.","Tailwind or Framer Motion is not already in the stack.","You need a full application component library."],"receipts":["https://magicui.design","https://github.com/magicuidesign/magicui"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"make","name":"Make","tagline":"Visual automation platform — connect 3,000+ apps without writing code.","url":"https://www.make.com","category":"automation","pricing":"freemium","price_note":"free 1,000 credits/mo (2 active scenarios, 15-min intervals); Core $12/mo billed yearly for 10k credits","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"The most polished visual automation platform around — 3,000+ integrations and a scenario editor non-technical people actually ship with. The move from operations to credits made monthly costs harder to predict, and busy flows get expensive at volume. If you don't need self-hosting or agent nodes, this is still the default pick.","limitations":["Free-plan credits expire monthly; annual paid credits expire after 12 months.","No self-hosting; your automations live on Make's cloud.","Credit pricing adds up quickly once flows run at volume."],"strengths":["Most polished visual automation: 3,000+ apps and a scenario editor non-engineers ship with.","Default pick when you do not need self-hosting or agent nodes."],"use_for":["Ops and growth automations across SaaS tools.","Teams that will live on a visual canvas, not n8n or code."],"skip_when":["You need to self-host the automations.","Credit burn at volume is already unpredictable.","Free-plan credits expiring monthly would strand the flow."],"receipts":["https://www.make.com","https://www.make.com/en/pricing"],"affiliate":"declared","affiliate_url":"https://www.make.com/en/register?pc=noemium","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"manus","name":"Manus","tagline":"General-purpose autonomous agent that executes multi-step web tasks.","url":"https://manus.im","category":"agents","pricing":"freemium","price_note":"free daily credits; Pro from $20/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"bundled","verdict":"skip","verdict_text":"Impressive demo reel, flaky day-to-day. Long-horizon tasks still fall apart, and failed runs burn credits anyway. Wait until autonomous execution is actually reliable.","limitations":["High failure rate on unscripted, real-world tasks.","Credits are consumed even on failed attempts.","Opaque about what it is doing mid-run."],"strengths":["Demo reel shows impressive autonomous multi-step web task execution.","Free daily credits let you try the agent before paying."],"use_for":["Benchmarking how well current autonomous agents handle long-horizon web tasks.","Quick experiments where a failed run costs only free daily credits."],"skip_when":["You need reliable execution on unscripted, real-world tasks.","Failed runs would waste paid credits or miss a deadline.","You require transparency about what the agent is doing mid-run."],"receipts":["https://manus.im","https://manus.im/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"redshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"marqo","name":"Marqo","tagline":"AI-native product discovery platform that trains dedicated models on catalogs and shopper behavior.","url":"https://www.marqo.ai","category":"data","pricing":"paid","free_tier":false,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Pricing and self-hosted options are not published; the platform is commerce-centric."],"receipts":["https://www.marqo.ai"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"mastra","name":"Mastra","tagline":"TypeScript agent framework plus a hosted platform — OSS to build, Cloud to observe and run.","url":"https://mastra.ai","category":"agents","pricing":"freemium","price_note":"framework OSS; Platform Starter $0 (100k obs events, 24 CPU-h); Teams $250/mo; gateway tokens at market + 5.5%","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"A TS-native LangGraph-shaped kit with evals and a real price page for the hosted control plane. Starter is $0 plus overage; Teams is $250. Gateway markup is 5.5% on tokens. Persistent 24/7 servers are $100/project extra. Framework ≠ platform bill.","limitations":["Hosted CPU, events, and tokens are three meters; a chatty agent trips all three.","5.5% gateway markup on top of provider rates if you use their gateway.","Persistent servers are a $100/project add-on, not included in Starter.","Self-host still needs you to run evals, tracing storage, and secrets."],"strengths":["TypeScript-native agent framework with evals built in.","Hosted control plane has a real, published price page.","Starter tier is $0 and self-host is available."],"use_for":["TypeScript teams building agents that need observability.","Teams that want a LangGraph-shaped kit without the Python stack."],"skip_when":["You cannot track three billing meters at once.","The 5.5% gateway markup is a dealbreaker.","Persistent 24/7 servers are needed but the $100 add-on is too much."],"receipts":["https://mastra.ai/pricing","https://github.com/mastra-ai/mastra"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"mem-ai","name":"Mem","tagline":"AI notes and knowledge base that organizes itself from your captures.","url":"https://get.mem.ai","category":"productivity","pricing":"freemium","price_note":"free tier; paid from $9/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Fast capture with AI search and automatic tagging. Good for people who want a second brain without manual organization. The graph view and exports are weaker than Obsidian or Notion.","limitations":["Export options are limited.","The AI tagging can miscategorize messy notes.","Less flexible for structured project management."],"strengths":["Fast capture with AI search and automatic tagging.","Low-friction second brain for people who avoid manual organization."],"use_for":["Personal note capture when search and auto-tags matter more than structure.","Quick retrieval from a messy stream of ideas and clips."],"skip_when":["You need robust export options or a graph view like Obsidian.","AI tagging miscategorizing messy notes would be costly to fix.","Structured project management is the main job."],"receipts":["https://get.mem.ai/pricing","https://docs.mem.ai/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"mem0","name":"Mem0","tagline":"Drop-in long-term memory layer for AI agents and apps — open-source library plus managed cloud platform.","url":"https://www.mem0.ai","category":"dev-infra","pricing":"freemium","price_note":"Open-source library and self-hosted server are free; managed Hobby plan is free (10k add requests/mo), Starter $19/mo, Pro $249/mo, Enterprise custom.","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"Mem0 gives agents persistent, queryable memory without rebuilding retrieval pipelines from scratch. The open-source Python/JS SDK and Docker-based self-hosted server make it easy to prototype, and the managed platform adds analytics and graph memory. It is still a young layer; production teams need to think hard about memory accuracy, privacy, and the cost of retrieval at scale.","limitations":["Memory quality depends on the underlying LLM and embedding model; bad extractions compound over time.","Free and Starter plans have tight request limits; production scale quickly lands on the $249/mo Pro tier or Enterprise.","Self-hosted setup requires managing vector storage and auth, not a one-line drop-in for most teams."],"strengths":["Drop-in long-term memory layer for agents and apps.","Open-source Python/JS SDK plus a Docker-based self-hosted server.","Managed platform adds analytics and graph memory on top of the core."],"use_for":["Agents and apps that need persistent, queryable memory without custom RAG.","Teams that want a self-hosted memory layer under their own infrastructure."],"skip_when":["Memory accuracy, privacy, and retrieval cost at scale are unproven concerns.","You cannot manage vector storage and auth for a self-hosted deployment.","Tight request limits on low tiers would break your production workload."],"receipts":["https://www.mem0.ai","https://www.mem0.ai/pricing","https://github.com/mem0ai/mem0"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"memory-mcp","name":"Memory MCP","tagline":"Official MCP memory server — a local entity graph the agent can read and write.","url":"https://github.com/modelcontextprotocol/servers/tree/main/src/memory","category":"mcp","pricing":"free","price_note":"MIT reference; npx -y @modelcontextprotocol/server-memory","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"situational","verdict_text":"A local knowledge graph (entities, relations, observations) on disk. Fine as the reference memory, and still the snippet tutorials copy. If you want a product on this job, that is Cogni, not this demo.","limitations":["Teaching graph, not a managed memory product.","Lives in the reference-servers repo; no support desk.","Effectiveness is whether the agent actually writes and recalls."],"strengths":["Official reference graph with entities, relations, and atomic observations.","Local, MIT, npx — no hosted memory leaving the machine."],"use_for":["A first persistent-memory MCP on one machine, one agent.","Demos of graph memory without a vendor account."],"skip_when":["You need hosted, multi-client memory — use Cogni.","The agent already has project memory (Claude Code, Cursor, Noemium).","You expected search quality beyond exact entity names."],"receipts":["https://github.com/modelcontextprotocol/servers/tree/main/src/memory"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"no","local_processing":true},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"merlin","name":"Merlin","tagline":"Browser AI assistant — 70+ tools and top models in one extension, on every page.","url":"https://www.getmerlin.in","category":"productivity","pricing":"freemium","price_note":"free 102 queries/day; Pro $19/mo; Teams $15/seat/mo (5-seat min, $180/yr per seat)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"An everything-in-one browser assistant: chat, writing, research and image tools behind a single subscription, available on any page. Convenient if you want one hub instead of a dozen tabs and logins. The bundle cuts both ways — \"unlimited\" runs on fair use, and a direct model subscription buys more control for similar money. Pick it for convenience, not for depth.","limitations":["\"Unlimited\" Pro queries are subject to a fair-use policy.","Free tier is capped at 102 queries per day.","Teams plan requires a 5-seat minimum."],"strengths":["Browser extension puts 70+ tools and top models on any page.","Single subscription replaces multiple model logins.","Free tier offers 102 queries/day to test the bundle."],"use_for":["Users who want one browser hub instead of juggling model tabs.","Quick writing, research, and image tasks inside the existing browser flow."],"skip_when":["\"Unlimited\" Pro fair-use caps would throttle your volume.","You need a single dedicated model with more control than the bundle allows.","You are buying for a team smaller than the 5-seat Teams minimum."],"receipts":["https://www.getmerlin.in/pricing"],"affiliate":"declared","affiliate_url":"https://www.getmerlin.in/chat?ref=zjfhyzh","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"midjourney","name":"Midjourney","tagline":"The aesthetic benchmark for AI image generation — Discord-born, now web-first.","url":"https://www.midjourney.com","category":"image","pricing":"paid","price_note":"from $10/mo (Basic)","free_tier":false,"open_source":false,"api":false,"self_host":false,"verdict":"ship","verdict_text":"Still the easiest route to visually striking images — light, texture and composition usually land on the first try. The web app fixed the Discord friction. No free tier, no real API, and images stay public unless you pay for stealth.","limitations":["No free tier at all.","No public API; automation requires unofficial workarounds.","Images are public unless you pay for a stealth-capable tier."],"strengths":["Easiest route to visually striking images — light, texture and composition usually land first try.","The web app removed the old Discord-only friction."],"use_for":["Concept art and marketing stills where aesthetics beat control.","A human-in-the-loop image desk, not an automated pipeline."],"skip_when":["You need a free tier to evaluate.","You need a public API.","Images must stay private and you will not pay for stealth."],"receipts":["https://docs.midjourney.com"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":true,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"milvus","name":"Milvus","tagline":"Open-source vector database for high-dimensional similarity search on massive datasets.","url":"https://milvus.io","category":"data","pricing":"freemium","price_note":"Milvus Lite is free for local prototyping; Zilliz Cloud offers paid serverless and dedicated cluster tiers.","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Distributed deployment and enterprise features carry operational complexity beyond the local library."],"receipts":["https://milvus.io"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"mindstudio","name":"MindStudio","tagline":"No-code AI agent platform with 200+ models and 1,000+ integrations, plus Remy for building apps from specs.","url":"https://www.mindstudio.ai","category":"agents","pricing":"freemium","price_note":"free tier: 1 agent, 1,000 runs/mo; Individual $20/mo ($16/mo annual); Business custom; AI model usage billed at cost on top","free_tier":true,"open_source":false,"api":true,"self_host":true,"model_routing":"bundled","verdict":"situational","verdict_text":"Fastest path from idea to deployed AI agent for non-technical teams. The 200+ model router and 1,000+ integrations remove API-key juggling, and Remy can scaffold a full app from a spec. The free tier is tiny, model costs are extra, and sophisticated non-AI workflow automation is better handled by n8n or Make.","limitations":["Free tier caps you at 1 agent and 1,000 runs/month.","AI model usage is charged at direct API rates on top of the subscription.","Self-hosting, SSO, and custom SLAs are Enterprise-only.","Less suited for pure, complex workflow automation than dedicated automation tools.","Vendor lock-in: agents, connectors, and managed DB live on the platform."],"strengths":["Fastest path from idea to deployed AI agent for non-technical teams.","200+ model router and 1,000+ integrations remove API-key juggling.","Remy can scaffold a full app from a spec.","Free tier lets you prototype before committing."],"use_for":["Non-technical teams that need a deployed AI agent quickly.","Projects that benefit from broad model choice and pre-built integrations."],"skip_when":["The free tier's 1 agent and 1,000 runs/month would throttle you.","You want model costs bundled into the subscription rather than billed at cost.","Sophisticated non-AI workflow automation is the real need.","You require self-hosting, SSO, or custom SLAs without an Enterprise contract."],"receipts":["https://www.mindstudio.ai","https://www.mindstudio.ai/pricing","https://github.com/mindstudio-ai/mindstudio-agent"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"minimax-api","name":"MiniMax API","tagline":"MiniMax developer API for text, image, video, music and speech models; also powers Hailuo AI video.","url":"https://platform.minimax.io/docs/guides/pricing-paygo","category":"models-api","pricing":"paid","price_note":"MiniMax-M3 ≤512K context $0.30/$1.20 per 1M (promo 50% off list $0.60/$2.40), >512K 2× rate; M2.5 $0.30/$1.20 per 1M; Text-01 not listed on current API docs","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["minimax-m3"],"model_routing":"locked","verdict":"situational","verdict_text":"M3 is cheap for high-volume text, and the same platform runs Hailuo video, speech and music. Just watch the pricing: list vs promo rates are unclear, and the open-weight flagship Text-01 has vanished from the main API docs.","limitations":["The hosted API is proprietary; any open-weight MiniMax models are separate releases.","Pricing page lists promo vs standard rates; promo expiration unclear.","MiniMax-Text-01, the open-weight flagship, appears deprecated or hidden from main API docs.","Western documentation and support are thinner than Qwen/DeepSeek.","Long-context input (>512K) is billed at 2× the standard rate."],"strengths":["Cheap high-volume text pricing on M3.","Same platform exposes video, speech, and music models (Hailuo).","Free tier exists for experimentation."],"use_for":["High-volume text generation where M3 pricing beats alternatives.","Projects that want video, speech, and music from one API vendor."],"skip_when":["You need clear, stable pricing without promo/standard confusion.","The open-weight Text-01 flagship is required for your stack.","Western documentation and support are non-negotiable."],"receipts":["https://platform.minimax.io/docs/guides/pricing-paygo","https://china-llm.com/blog/minimax-api-pricing","https://developer.puter.com/tutorials/minimax-api-pricing/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"minimax-speech","name":"MiniMax Speech","tagline":"Text-to-speech, voice cloning, and audio generation API.","url":"https://platform.minimax.io/docs/guides/pricing-paygo","category":"audio","pricing":"paid","price_note":"Turbo $60/M characters, HD $100/M characters; subscription plans $5–$150/month on some aggregators; voice clone $1.5/clone on fal.ai","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Strong Chinese voice quality and expressive speech at budget prices. Character-based pricing can get expensive at scale, language coverage is narrower than ElevenLabs, and Western access usually goes through resellers.","limitations":["Character-based pricing can be expensive at scale vs token-based TTS.","Less language coverage than ElevenLabs or Azure.","Western access is reseller-dependent.","Documentation and brand recognition outside China are limited."],"strengths":["Strong Chinese voice quality and expressive speech.","Budget character-based pricing.","Voice cloning available via fal.ai."],"use_for":["Chinese TTS and voice-clone projects.","Budget speech generation where character pricing fits."],"skip_when":["You need broad Western language coverage.","Token-based pricing is cheaper at your scale.","Direct Western access without resellers is required."],"receipts":["https://platform.minimax.io/docs/guides/pricing-paygo","https://knowara.com/ai-tools/voice/minimax-speech-review/","https://fal.ai/models/fal-ai/minimax/voice-clone"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"mistral-api","name":"Mistral API (La Plateforme)","tagline":"European frontier lab — efficient models, open-weights options, EU hosting.","url":"https://docs.mistral.ai","category":"models-api","pricing":"freemium","price_note":"free plan incl. $10/mo API credits; pay-per-token (Large 3: $0.50/$1.50 per 1M)","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["mistral-large-2512"],"model_routing":"locked","verdict":"situational","verdict_text":"The pick when EU data residency or open weights are non-negotiable. The small models are excellent value. For raw frontier quality, Claude and GPT still lead, and the product naming keeps changing.","limitations":["Frontier quality a half-step behind OpenAI/Anthropic.","Thinner ecosystem and docs.","Product naming and tiers are confusing.","Le Chat renamed to Vibe, console is now Mistral Studio (2026 rebrand)."],"strengths":["EU data residency option.","Open-weights options available.","Small models are excellent value."],"use_for":["Projects where EU data residency is required.","Teams wanting open-weights options from a frontier lab.","Cost-sensitive inference on Mistral's small models."],"skip_when":["Raw frontier quality must beat Claude/GPT.","You need a mature ecosystem and docs.","Confusing product renames would break internal documentation."],"receipts":["https://docs.mistral.ai","https://docs.mistral.ai/getting-started/models/models_overview"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"mistral-vibe","name":"Mistral Vibe","tagline":"Mistral's assistant (ex Le Chat) plus Vibe coding CLI/IDE — not the La Plateforme API card.","url":"https://mistral.ai/products/vibe/","category":"productivity","pricing":"freemium","price_note":"Free $0; Pro $14.99/mo (students $5.99); Team $24.99/user/mo; $30/mo API credits on Pro; usage beyond plan at API rates","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["mistral-large-2512"],"verdict":"situational","verdict_text":"Le Chat was renamed Vibe. Free is a real assistant; Pro at $14.99 is the all-day coding and long-horizon SKU, cheaper than ChatGPT Plus on the public page. API (La Plateforme) remains a different bill. EU residency story is the procurement win. Not field-tested as a daily driver here.","limitations":["Vibe Pro and API spend are separate meters; CLI can blow past the sub.","Free coding sessions are limited; Pro is the CLI/IDE unlock.","Connectors and MCP are beta.","Not a duplicate of the Mistral API card."],"strengths":["Pro at $14.99 is cheaper than ChatGPT Plus on the public page.","Free tier is a real assistant, not just a teaser.","EU residency and procurement story is the headline advantage."],"use_for":["Users who prioritize EU data residency and vendor location.","All-day coding and long-horizon tasks where Pro unlocks the CLI/IDE."],"skip_when":["You need assistant and API spend on a single bill.","You rely on mature connectors or MCP that are still in beta.","Free coding sessions are enough for your workflow."],"receipts":["https://mistral.ai/pricing/","https://mistral.ai/products/vibe/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"mobbin","name":"Mobbin","tagline":"Screenshots and flow library of real mobile and web app designs.","url":"https://mobbin.com","category":"design","pricing":"freemium","price_note":"free limited browse; Pro kr100/mo billed yearly","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"ship","verdict_text":"Essential reference when you want to see how top apps solve the same UI problem. The library is huge and the tagging makes it fast to find patterns.","limitations":["Paid plans are needed for full access and downloads.","Focus is on screenshots, not code.","Can encourage imitation over original thinking."],"strengths":["Huge tagged library of real app screens — how top products solved the same UI problem.","Fast reference before you invent a flow that already exists."],"use_for":["Product and design research on mobile and web patterns.","Reviews where you need screenshots, not a component kit."],"skip_when":["You need downloads and full access on the free browse.","The deliverable is code, not reference.","The team would copy flows instead of deciding."],"receipts":["https://mobbin.com","https://mobbin.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"modal","name":"Modal","tagline":"Serverless GPU and compute platform for running ML and AI workloads.","url":"https://modal.com","category":"dev-infra","pricing":"freemium","price_note":"free tier ($30/mo compute); pay-as-you-go beyond","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"The easiest way to deploy a Python function onto a GPU without touching Kubernetes. Excellent for fine-tuning, inference, and batch jobs. Pricing is usage-based, so runaway jobs can get expensive fast.","limitations":["Python-first; other languages have limited support.","Usage-based GPU costs require cost monitoring.","Cold starts can be noticeable for some workloads."],"strengths":["Easiest way to put a Python function on a GPU without touching Kubernetes.","Fits fine-tunes, inference and batch jobs that should not own a cluster."],"use_for":["Python ML jobs that spike and then go quiet.","Teams that want serverless GPUs more than a long-lived box."],"skip_when":["The runtime is not Python.","Runaway GPU jobs would go unmonitored.","Cold starts would show up in the user-facing path."],"receipts":["https://modal.com","https://modal.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"modelmap","name":"modelmap","tagline":"Paste a Hugging Face model ID and get a living, animated architecture map without downloading weights.","url":"https://modelmap.cc/","category":"dev-infra","pricing":"free","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Instantly visualizes transformer blocks, attention, MoE routing, and shapes for any public HF model. Useful for researchers and practitioners who need to understand a new architecture before they commit to a download. No code or GPU required.","limitations":["Only works for public Hugging Face repos; gated models need a token.","Visualization fidelity depends on what the traced fake forward pass can infer.","No evidence of an export or embedding API yet."],"strengths":["Instantly visualizes transformer blocks, attention, MoE routing, and shapes.","No code or GPU needed to inspect a model architecture.","Works for any public Hugging Face model."],"use_for":["Researchers understanding a new architecture before downloading weights.","Practitioners who need a quick visual map of a model's structure."],"skip_when":["You need to inspect gated models without a Hugging Face token.","You need an export or embedding API for the visualization.","The traced fake forward pass cannot infer the details you care about."],"receipts":["https://modelmap.cc/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"morgen","name":"Morgen","tagline":"AI-powered daily planner that combines calendars and tasks from multiple apps into one schedule.","url":"https://morgen.so","category":"productivity","pricing":"paid","price_note":"Free trial available; paid subscription required after trial.","free_tier":false,"open_source":false,"api":false,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Morgen Assist currently integrates with Outlook, Google, iCloud, and Fastmail calendars only, and some organizations limit third-party calendar tools."],"receipts":["https://morgen.so","https://morgen.so/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"motion","name":"Motion (usemotion)","tagline":"AI project manager that auto-schedules tasks into your calendar.","url":"https://www.usemotion.com","category":"productivity","pricing":"paid","price_note":"7-day trial; Pro AI $19/seat/mo yearly (7,500 credits/seat); Business AI $29/seat/mo yearly (15,000 credits/seat)","free_tier":false,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Auto-scheduling works when your work is a pile of deadline-driven tasks. It is opinionated, the setup takes time, and the price adds up fast. Try it if your calendar is the problem; skip it if you just want a simple to-do list.","limitations":["Expensive for a personal productivity tool.","Requires aggressive onboarding to get value.","Notifications can feel like a micromanaging boss.","No free tier; AI credits model (7.5k credits/seat on Pro)."],"strengths":["Auto-schedules deadline-driven tasks directly into your calendar.","Yearly plans include a predictable AI-credit pool per seat."],"use_for":["People whose calendar is the bottleneck and tasks are deadline-driven.","Teams with a calendar-centric workflow and budget for a productivity tool."],"skip_when":["You just want a simple to-do list without calendar rearrangement.","The price and onboarding overhead outweigh the scheduling benefit.","AI-credit limits would throttle how you plan work."],"receipts":["https://www.usemotion.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"motion-dev","name":"Motion","tagline":"JavaScript and React animation library, formerly Framer Motion.","url":"https://motion.dev","category":"design","pricing":"free","price_note":"MIT open source","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"ship","verdict_text":"The default choice for React animations for good reason. The API is expressive, the performance is solid, and the new Motion One variant covers vanilla JS projects too.","limitations":["Bundle size can add up if you import everything.","Layout animations still have edge cases.","The rebrand from Framer Motion created some documentation churn."],"strengths":["Default React animation library: expressive API, solid performance.","Motion One covers vanilla JS when the project is not React."],"use_for":["Product UI motion in React.","A shared animation layer across React and a little vanilla JS."],"skip_when":["Importing the whole library would bloat the bundle.","Layout animations hit known edge cases in your UI.","You are still lost in the Framer Motion rebrand docs."],"receipts":["https://motion.dev","https://github.com/motiondivision/motion"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"murf","name":"Murf AI","tagline":"Business-focused voiceover studio for presentations and e-learning.","url":"https://murf.ai","category":"audio","pricing":"freemium","price_note":"free tier; Creator $29/mo or $19/mo billed yearly","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"A safe pick for corporate training and slide voiceovers where consistency and team licenses matter more than sounding truly human. APIs and consumer cloning are not its strength — ElevenLabs is the better fit there. Good enough for e-learning, just don't expect a warm, natural voice.","limitations":["Voices sound more corporate-synthetic than the leaders.","Limited voice-cloning flexibility.","Editor-centric product; weaker API story."],"strengths":["Corporate-friendly voiceover studio with team licenses and export controls.","Consistency-first output fits training and presentation workflows."],"use_for":["Corporate training modules and e-learning narration.","Slide voiceovers where a steady, business-safe tone is enough."],"skip_when":["You need a warm, natural-sounding voice rather than a corporate-synthetic one.","Flexible voice cloning or a strong API is the main requirement.","ElevenLabs-level expressiveness is the benchmark."],"receipts":["https://murf.ai","https://murf.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"n8n","name":"n8n","tagline":"Source-available workflow automation with strong AI agent nodes.","url":"https://n8n.io","category":"automation","pricing":"freemium","price_note":"self-host free; cloud Starter €24/mo (€20/mo billed yearly)","free_tier":true,"open_source":false,"api":true,"self_host":true,"verdict":"ship","verdict_text":"Visual workflows with real AI agent nodes, plus you can self-host. That's a rare mix of approachable and controllable. Fair-code, not fully open source, and big workflows get messy to debug. Still the strongest Zapier alternative once AI steps enter the picture.","limitations":["Fair-code license, not true open source for resale.","Error handling in big workflows gets messy.","Cloud per-execution pricing stings at volume."],"strengths":["Visual workflows with real AI agent nodes, and you can self-host.","Strongest Zapier-class alternative once AI steps enter the picture."],"use_for":["Self-hosted automations that need agent nodes, not just HTTP.","Ops teams that outgrew Make and want the workflows on their metal."],"skip_when":["Fair-code licensing blocks the way you would resell it.","Big workflows are already painful to debug.","Cloud per-execution pricing would sting at your volume."],"receipts":["https://github.com/n8n-io/n8n","https://n8n.io/pricing"],"affiliate":"none","evidence_tier":"source-verified","guide":{"install":"Self-host with Docker (n8n docs) or n8n Cloud. Pin the image tag.","requirements":["A place to run Docker or a Cloud workspace","Credentials for every node you enable"],"cost":"Self-host software is free; Cloud starts around €20–24/mo on the public list. Execution volume is the real meter.","security":"Every credential in the workflow is a secret. Do not expose the editor to the internet without auth.","breaks_when":["A workflow has dozens of branches and no tests — debug time exceeds build time.","Fair-code licensing blocks the way you wanted to resell it."]},"momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"navbar-gallery","name":"Navbar Gallery","tagline":"Curated gallery of navigation bar designs for web projects.","url":"https://navbar.gallery","category":"design","pricing":"free","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Curated navbar screenshots for early design reference, not copy-paste components. Useful for locking a navigation direction before you open Figma, but expect to build and refine the markup yourself and do not count on a steady stream of new examples.","limitations":["No code export; just screenshots and ideas.","Search and filtering are basic.","Updates depend on the curator's pace."],"strengths":["Curated navigation patterns from real websites, sorted by style.","Quick visual reference before you design your own navbar."],"use_for":["Early ideation for a navigation component.","Mood-boarding header directions with a team."],"skip_when":["You need exported code; only screenshots and ideas are shown.","You need powerful search and filtering beyond the basics.","The curator's update pace is too slow for your project timeline."],"receipts":["https://navbar.gallery","https://www.navbar.gallery/style/minimal"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"nebius","name":"Nebius","tagline":"AI cloud platform with GPU VMs and API access to open models.","url":"https://nebius.com","category":"models-api","pricing":"paid","price_note":"pay-as-you-go; from $0.40/1M tokens for smaller models","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Aggressively priced inference and GPU cloud, useful for cost-sensitive workloads. The platform is newer, so documentation and ecosystem are still catching up to bigger providers.","limitations":["Newer platform with a smaller ecosystem.","Documentation and support are less mature.","Enterprise features still expanding."],"strengths":["Aggressively priced inference and GPU cloud for cost-sensitive workloads.","API access to open models without running your own GPUs.","Unified AI cloud platform with compute and inference in one account."],"use_for":["Cost-sensitive inference that wants open models without self-hosting.","GPU cloud jobs where you rent by the hour instead of owning hardware."],"skip_when":["You need mature documentation and support from a bigger provider.","Enterprise features and ecosystem breadth are critical now.","You cannot accept a newer platform still catching up."],"receipts":["https://nebius.com","https://docs.nebius.com/studio/inference/models/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"notebooklm","name":"NotebookLM","tagline":"Google's research notebook — grounded Q&A and audio overviews over your own sources.","url":"https://notebook.google/","category":"data","pricing":"freemium","price_note":"free; Plus via Google AI Plus $4.99/mo; Pro via Google AI Pro $19.99/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["gemini-3-1-pro-preview"],"verdict":"ship","verdict_text":"The most reliable consumer RAG for grounded reading: answers stick to your uploaded sources and cite them, and the audio overviews are genuinely useful for digesting material. The free tier is generous.","limitations":["Source count limits per notebook.","No API, so no automation.","Google product-graveyard risk is real."],"strengths":["Most reliable consumer RAG: answers stick to uploaded sources and cite them.","Audio overviews are actually useful for digesting a pile of PDFs."],"use_for":["Grounded reading over a known source set.","A research notebook when you do not need an API."],"skip_when":["You hit per-notebook source limits.","You need to automate it — there is no API.","Google product-graveyard risk is unacceptable."],"receipts":["https://notebook.google/","https://workspace.google.com/products/notebooklm/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"notion","name":"Notion","tagline":"AI workspace where teams capture context, find answers, and automate tasks together.","url":"https://www.notion.so","category":"productivity","pricing":"freemium","price_note":"Free for individuals with unlimited blocks; paid plans add AI features, guests, and advanced admin controls.","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Free plan blocks are unlimited for individuals but limited for multi-member workspaces, and Notion AI is only a limited trial on lower tiers."],"receipts":["https://www.notion.so","https://www.notion.so/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"no","local_processing":false},"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"notion-ai","name":"Notion AI","tagline":"AI assistant embedded in Notion — summarize, draft and search your workspace.","url":"https://www.notion.com/product/ai","category":"writing","pricing":"paid","price_note":"limited AI trial on free/Plus; full AI in Business $20/member/mo (yearly)","free_tier":false,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Makes sense only if your team already lives in Notion. Searching across your own workspace docs is the real feature; the writing assistant itself is fine, not special. Don't buy Notion just for the AI.","limitations":["Only makes sense if you already pay for Notion.","Per-seat pricing stacks on top of the base plan.","Which models it routes to is opaque.","Standalone $10 AI add-on discontinued."],"strengths":["Searching across your own workspace docs is the real feature.","Embedded where your notes, wikis and projects already live."],"use_for":["Teams already living in Notion who need workspace-wide AI search.","Summarizing and drafting pages that are already in Notion."],"skip_when":["You do not already pay for Notion as a workspace.","Transparent model routing and choice are required.","Per-seat AI pricing would stack too high on top of the base plan."],"receipts":["https://www.notion.com/product/ai","https://www.notion.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"notion-calendar","name":"Notion Calendar","tagline":"Calendar built on top of Notion, syncing meetings, tasks, and docs.","url":"https://www.notion.so/product/calendar","category":"productivity","pricing":"free","price_note":"free; requires Notion account","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Clean, minimalist calendar tightly linked to Notion pages and tasks. Great for Notion power users; a thin standalone calendar if you don't use Notion for project management.","limitations":["Deeply tied to the Notion ecosystem.","Limited advanced scheduling features compared to Calendly or Reclaim.","Mobile experience is simpler than desktop."],"strengths":["Clean, minimalist calendar tightly linked to Notion pages and tasks.","Free for anyone with a Notion account.","Syncs meetings, tasks and docs in one connected workspace."],"use_for":["Notion power users who already manage projects and docs in Notion.","Teams that want a calendar integrated with their existing Notion workspace."],"skip_when":["You do not use Notion for project management or task tracking.","You need advanced scheduling features like Calendly or Reclaim offer.","A full-featured mobile calendar experience is required."],"receipts":["https://www.notion.so/product/calendar","https://www.notion.so/help/notion-calendar-apps"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"notion-mcp","name":"Notion MCP","tagline":"Official Notion MCP server — workspace pages and databases as tools for coding agents.","url":"https://github.com/makenotion/notion-mcp-server","category":"mcp","pricing":"free","price_note":"Server is free; you need a Notion workspace and integration token","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: OAuth/token scopes not reproduced in this wave.","A writable Notion integration is a data-exfil path. Least privilege or skip."],"receipts":["https://github.com/makenotion/notion-mcp-server","https://developers.notion.com/docs/mcp"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"novita-ai","name":"Novita AI","tagline":"AI-native cloud with serverless APIs for 200+ text, image, audio, and video models plus dedicated endpoints.","url":"https://novita.ai","category":"models-api","pricing":"paid","price_note":"Usage-based per-token, per-image, and per-second rates for 200+ models; dedicated endpoints priced separately.","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Pricing varies widely by modality and model; video and audio workloads can exceed text-only costs."],"receipts":["https://novita.ai","https://novita.ai/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"nvidia-riva","name":"NVIDIA Riva","tagline":"GPU-accelerated SDK for customizable real-time speech recognition, synthesis, and translation pipelines.","url":"https://www.nvidia.com/en-us/ai-data-science/products/riva/","category":"audio","pricing":"paid","price_note":"Can be explored through managed endpoints on nvidia.build.com; production deployment is self-hosted on NVIDIA GPU infrastructure.","free_tier":false,"open_source":false,"api":true,"self_host":true,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Requires NVIDIA GPU infrastructure and is delivered as containerized microservices rather than a managed API."],"receipts":["https://www.nvidia.com/en-us/ai-data-science/products/riva/"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"obsidian","name":"Obsidian","tagline":"Private, local-first notes app with open file formats and thousands of plugins to shape your thinking.","url":"https://obsidian.md","category":"productivity","pricing":"freemium","price_note":"Core app is free for personal use; paid add-ons include Obsidian Sync, Publish, Catalyst, and commercial licenses.","free_tier":true,"open_source":false,"api":false,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Core app is free for personal use; organizations are encouraged to purchase a commercial license, and Sync/Publish/Catalyst are paid add-ons."],"receipts":["https://obsidian.md","https://obsidian.md/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"no","local_processing":true},"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"ollama","name":"Ollama","tagline":"Run open models locally (CLI, app, API) — optional Ollama Cloud Pro/Max for hosted large models.","url":"https://ollama.com","category":"dev-infra","pricing":"freemium","price_note":"local runtime $0 unlimited on your hardware; Cloud Free light usage; Pro $20/mo ($200/yr); Max $100/mo (new Max sign-ups paused)","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"The local-model default: one CLI, an OpenAI-shaped API, desktop apps, and you keep weights on disk. Cloud Pro is a separate $20 meter with 5-hour and weekly caps; Max is capacity-paused for new buyers. Local is the product; cloud is overflow. Not a quality claim about any one model.","limitations":["Local quality and speed are your GPU; Ollama is the runner, not the weights.","Cloud usage is token-and-model-weighted, not a flat token bucket.","New Max subscriptions paused while they add capacity (existing Max kept).","Cloud hosts primarily in the US (EU/Singapore overflow); data-residency still a question."],"strengths":["Local-model default with one CLI, an OpenAI-shaped API, and desktop apps.","You keep model weights on disk and run inference on your own hardware.","Free local runtime with no usage caps beyond what your machine can handle."],"use_for":["Offline or private inference where data never leaves the machine.","Prototyping with open weights before deciding on cloud spend."],"skip_when":["Your GPU cannot deliver the quality or speed the use case demands.","You need guaranteed data residency from a cloud region.","You were planning to sign up for Max while new subscriptions are paused."],"receipts":["https://ollama.com/pricing","https://github.com/ollama/ollama"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"onlook","name":"Onlook","tagline":"Open-source visual editor that writes code directly into your React project.","url":"https://onlook.com","category":"design","pricing":"freemium","price_note":"free open-source; paid cloud collaboration","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"situational","verdict_text":"Promising for teams that want a designer-friendly layer on top of a real codebase. It is still early, so expect rough edges and occasional sync issues.","limitations":["Early-stage; breaking changes are common.","Works best with specific React setups.","Cloud collaboration features are not essential for solo use."],"strengths":["Open-source visual editor that writes code directly into your React project.","Designer-friendly layer on top of a real codebase."],"use_for":["React teams that want a visual editor synced to a real codebase.","Designers and developers who share the same React repository."],"skip_when":["You need a stable, mature tool without breaking changes.","Your stack is not one of the React setups Onlook supports well.","Solo users who don't need the paid cloud collaboration tier."],"receipts":["https://onlook.com","https://github.com/onlook-dev/onlook"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"open-interpreter","name":"Open Interpreter","tagline":"Natural-language computer use: a ChatGPT-like prompt that runs code locally.","url":"https://openinterpreter.com","category":"coding","pricing":"freemium","price_note":"OSS local; optional hosted/computer-use products","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Full local computer use is a critical privilege boundary.","Radar: we have not reproduced current install against this year's CLI."],"receipts":["https://openinterpreter.com","https://github.com/openinterpreter/open-interpreter"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"open-webui","name":"Open WebUI","tagline":"Self-hosted chat UI for Ollama and OpenAI-compatible APIs — your models, your box.","url":"https://openwebui.com","category":"dev-infra","pricing":"freemium","price_note":"self-host free (pip/docker); Enterprise SSO/RBAC/air-gap is a commercial SKU (talk to sales)","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"The default homelab ChatGPT skin: Ollama, OpenAI, Anthropic, RAG, voice, a huge community library. The product is the interface, not the model. Enterprise exists for SSO and air-gap. Ops (auth, updates, disk) are on you. Not a cloud assistant.","limitations":["You operate Docker/GPU/backups; this is not a hosted chat product.","Feature surface is huge; upgrades can reshuffle the UI (see v0.11 notes).","Connecting cloud APIs reintroduces the data-residency question you self-hosted to avoid.","Enterprise controls are not the MIT community build."],"strengths":["Self-hosted chat UI that works with Ollama and OpenAI-compatible APIs.","Huge community library of models, tools and prompts.","RAG and voice features out of the box."],"use_for":["Homelabs that want a ChatGPT-like skin on local models.","Teams that need an air-gapped or self-hosted chat interface.","Users who already run Ollama and want a richer web UI."],"skip_when":["You want a fully managed cloud assistant with zero ops.","You cannot maintain Docker, GPU, backups and auth yourself.","You are self-hosting to avoid cloud APIs but will still plug them in."],"receipts":["https://github.com/open-webui/open-webui","https://docs.openwebui.com"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"openai-agents-sdk","name":"OpenAI Agents SDK","tagline":"Official Python agents runtime — agents, handoffs, guardrails, sessions; not the Responses API itself.","url":"https://openai.github.io/openai-agents-python/","category":"agents","pricing":"free","price_note":"SDK free (pip install openai-agents); model/tool calls billed at OpenAI API rates; tracing uses your OpenAI account","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"locked","verdict":"situational","verdict_text":"The runtime on top of Responses, not a second Responses card. Tiny primitives (agent, handoff, guardrail, session) plus sandbox and realtime paths. You own orchestration less than with raw Responses, more than with a visual builder. Tokens still land on the API bill.","limitations":["Default loop talks to OpenAI; other providers are possible but second-class.","Built-in tracing is convenient and also a data-boundary choice.","Not a no-code product; you write Python.","Swarm is the old experiment — this SDK is the replacement."],"strengths":["Official Python agents runtime with tiny, composable primitives.","Agent, handoff, guardrail, and session abstractions out of the box.","Includes sandbox and realtime paths beyond basic chat loops."],"use_for":["Python developers building agents on top of the OpenAI Responses API.","Teams wanting more orchestration than a no-code builder but less than raw API code."],"skip_when":["You need a visual no-code interface for building agents.","First-class non-OpenAI providers are a hard requirement.","Built-in tracing sending data to OpenAI is a data-boundary problem."],"receipts":["https://openai.github.io/openai-agents-python/","https://github.com/openai/openai-agents-python"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"openai-api","name":"OpenAI API","tagline":"GPT-5 family and o-series reasoning models behind one API.","url":"https://platform.openai.com","category":"models-api","pricing":"paid","price_note":"pay-as-you-go, from $0.20/1M input (GPT-5.6 Luna); GPT-5.6 Sol promotional $4/$20 at least through 2026-11-21 (list $5/$30); no free tier","free_tier":false,"open_source":false,"api":true,"self_host":false,"models_used":["gpt-5-6-terra","gpt-5-4-mini","gpt-5-6-sol","gpt-5-6-luna"],"model_routing":"locked","verdict":"ship","verdict_text":"The safe default for production: a model for every tier from nano to frontier, plus the most complete tooling — batch, structured outputs, agents. Rarely the cheapest, rarely the wrong choice. Just expect model names to shift on you.","limitations":["Premium pricing at the frontier tier.","Model deprecations move fast.","Conservative rate limits on new accounts."],"strengths":["Safe production default: a model for every tier plus batch, structured outputs and agents.","Rarely the cheapest, rarely the wrong choice for a first production API."],"use_for":["Shipping an app on the GPT-5 family without assembling a router.","Workloads that need OpenAI's tooling more than a discount host."],"skip_when":["Frontier-tier pricing is the only quality that works.","You cannot absorb fast model deprecations.","New-account rate limits would stall launch."],"receipts":["https://platform.openai.com/docs"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"openai-codex","name":"OpenAI Codex","tagline":"OpenAI's coding agent — terminal CLI, IDE, and cloud tasks, bundled into ChatGPT plans.","url":"https://chatgpt.com/codex","category":"coding","pricing":"freemium","price_note":"included in ChatGPT Free/Go/Plus/Pro; Plus $20/mo, Pro 5x $100/mo or 20x $200/mo; extra credits after 5h caps; API key billed at API rates","free_tier":true,"open_source":true,"api":true,"self_host":false,"models_used":["gpt-5-6-sol","gpt-5-6-terra","gpt-5-6-luna"],"model_routing":"bundled","verdict":"situational","verdict_text":"The CLI on GitHub is real and the ChatGPT bundle makes it the lowest-friction OpenAI coding agent. Caps sit on a five-hour window, credits replaced per-message billing in Apr 2026, and cloud tasks are Plus-and-up. Not a second Responses API card — this is the product you run against a repo. No field run here, so this is not a Ship stamp.","limitations":["Local and cloud usage share a rolling 5-hour window; weekly caps also apply.","Free/Go are trial-level; cloud GitHub review and plugins sit on Plus+.","Token-credit metering (from 2026-04) makes a session's cost hard to predict.","Conversation/repo context still leaves OpenAI; no self-host of the hosted agent."],"strengths":["Lowest-friction OpenAI coding agent if you already pay for ChatGPT.","Real CLI on GitHub plus IDE and cloud tasks on one bundle.","Free/Go exist as a trial path; Plus is where cloud GitHub review lives."],"use_for":["Repo work on an OpenAI seat you already have, not a second Responses API card.","Mixed local CLI and cloud tasks inside the ChatGPT plan."],"skip_when":["You need a predictable session cost — credits and a 5-hour window fight you.","The repo cannot leave OpenAI's hosted agent.","Multi-step agent work is the job and you can run Cursor or Claude Code instead."],"receipts":["https://chatgpt.com/codex/pricing/","https://github.com/openai/codex","https://help.openai.com/en/articles/20001106-codex-rate-card"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"openai-responses","name":"OpenAI Responses API","tagline":"OpenAI's agent-building API — tools, web search and file search built in.","url":"https://platform.openai.com/docs/api-reference/responses","category":"agents","pricing":"paid","price_note":"pay-per-token; e.g. GPT-5.6 Luna $0.20/1M input","free_tier":false,"open_source":false,"api":true,"self_host":false,"models_used":["gpt-5-6-terra","gpt-5-4-mini","gpt-5-6-sol","gpt-5-6-luna"],"model_routing":"locked","verdict":"ship","verdict_text":"Hosted web search, file search, and code interpreter without building plumbing — the most batteries-included agent API. Conversation state lives on OpenAI's side, which keeps things simple until it doesn't. A sensible default if you want an agent running today instead of wiring your own graph.","limitations":["Hosted-tool pricing (storage, search calls) sneaks up on you.","Conversation state lives on OpenAI's side.","Less controllable than explicit-graph frameworks like LangGraph.","Assistants API sunsets 2026-08-26 — Responses is the default for new projects."],"strengths":["Batteries-included agent API: hosted web search, file search and code interpreter.","Sensible default when you want an agent running today, not a graph framework."],"use_for":["New OpenAI agent projects after the Assistants sunset.","Prototypes that should use hosted tools instead of your own retrieval."],"skip_when":["Hosted-tool storage and search pricing would sneak up.","Conversation state cannot live on OpenAI's side.","You need LangGraph-level control of the graph."],"receipts":["https://platform.openai.com/docs/api-reference/responses","https://platform.openai.com/docs/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"opencode","name":"OpenCode","tagline":"Autonomous coding agent that edits files, runs tests, and validates changes in your repo.","url":"https://opencode.ai","category":"coding","pricing":"freemium","price_note":"free tier; paid plans from $20/mo","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"ship","verdict_text":"Spins up a full agent loop in your project, writes code, runs tests, and commits when it succeeds. Good for well-scoped tasks with existing test coverage. Still rough on large refactors and can hallucinate library APIs.","limitations":["Agent loops can wander on ambiguous requirements.","Token consumption on large codebases adds up fast.","Less mature editor integration than Cursor or Windsurf."],"strengths":["Full agent loop in your repo: edit, test, commit when it succeeds.","Open-source path for well-scoped tasks that already have tests."],"use_for":["Bounded coding jobs with existing test coverage.","A terminal agent when you do not want a full AI IDE."],"skip_when":["The task is an ambiguous large refactor.","Token burn on a big repo would dominate the bill.","You need mature editor integration like Cursor."],"receipts":["https://opencode.ai","https://github.com/anomalyco/opencode"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"openhands","name":"OpenHands","tagline":"Open-source software-engineering agent — GUI, CLI, SDK, plus a hosted Cloud.","url":"https://openhands.dev","category":"coding","pricing":"freemium","price_note":"self-host MIT/community repos free (BYOK); Cloud is commercial at app.all-hands.dev — no public HTML $ card fetched, confirm in-product","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"The open Devin-shaped stack: Agent Canvas in the browser, a Python SDK, CLI, and sandboxes you can run yourself. Cloud is a hosted SKU, not the license. Repos split MIT vs commercial enterprise directories — read the LICENSE in the repo you actually deploy. Not field-run here.","limitations":["Cloud pricing is in-product; this card will not invent a $20 figure from blogs.","Multiple repos (Canvas, SDK, sandbox, automation) — \"OpenHands\" is a family.","You still bring a model; sandbox compute is the other bill.","Legacy Local GUI is deprecated; Canvas is the current browser client."],"strengths":["Open-source Devin-shaped stack with browser Canvas, Python SDK, CLI, and sandboxes.","Self-hostable sandboxes keep the agent under your control."],"use_for":["Teams that want an open, auditable coding agent they can run themselves.","Building on the Canvas browser client rather than the deprecated Local GUI."],"skip_when":["You need public Cloud pricing before you can evaluate cost.","You are not prepared to bring your own model and pay for sandbox compute.","You cannot navigate the split between MIT community and commercial enterprise repos."],"receipts":["https://docs.openhands.dev","https://github.com/OpenHands/OpenHands","https://docs.openhands.dev/openhands/usage/cloud/openhands-cloud"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"openrouter","name":"OpenRouter","tagline":"One API for every major model — unified billing, fallbacks and routing.","url":"https://openrouter.ai","category":"dev-infra","pricing":"freemium","price_note":"pay-per-token pass-through plus a small fee","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["gpt-5-6-terra","claude-sonnet-5","gemini-3-1-pro-preview","deepseek-v4-flash","qwen3-8-max","llama-4-maverick","mistral-large-2512","grok-4-6"],"model_routing":"locked","verdict":"ship","verdict_text":"Small teams get a lot of mileage out of this. Change one string to swap models, see all costs on one bill, get fallbacks when a provider flakes. You pay a small markup for the convenience, but it saves you from wiring a dozen APIs yourself.","limitations":["Small markup over direct API pricing.","Adds a network hop and latency.","Provider-specific features arrive later than direct APIs."],"strengths":["One string to swap models, one bill, fallbacks when a provider flakes.","Small teams avoid wiring a dozen lab APIs themselves."],"use_for":["Apps that should route across labs without a custom gateway.","Prototypes that will change models more than once a week."],"skip_when":["The markup over direct pricing is not worth the convenience.","An extra network hop would miss a latency budget.","You need provider-specific features the day they ship."],"receipts":["https://openrouter.ai","https://openrouter.ai/docs"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":true,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"origin-ui","name":"COSS UI (formerly Origin UI)","tagline":"Open-source copy-paste React components, evolved from Origin UI.","url":"https://coss.com/ui","category":"design","pricing":"free","price_note":"free and open source","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Origin UI has moved into COSS UI after its acquisition by Cal.com. The new library remains a polished shadcn-compatible alternative with a distinct visual style, but the migration is still in progress.","limitations":["Smaller ecosystem than shadcn/ui.","Original Origin UI components are now legacy and receive limited support.","The COSS UI roadmap is still in an active migration phase.","Style is opinionated; may not fit every brand."],"strengths":["Polished shadcn-compatible alternative with a distinct visual style.","Free and open source after the Cal.com acquisition.","Components remain usable while the COSS UI migration proceeds."],"use_for":["Teams that want a shadcn alternative with a different visual direction.","Projects that can tolerate a migration-phase component library."],"skip_when":["You need the ecosystem maturity and community of shadcn/ui today.","The opinionated style does not match the brand you are building.","You cannot accept roadmap uncertainty during the COSS UI migration."],"receipts":["https://coss.com/ui/docs/roadmap","https://github.com/cosscom/coss"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"otter-ai","name":"Otter.ai","tagline":"Meeting transcription and summaries with deep calendar integration.","url":"https://otter.ai","category":"productivity","pricing":"freemium","price_note":"Basic free (300 min/mo); Business $19.99/user/mo (annual)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Solid transcription and decent search across meeting history — the safe pick for teams that already use it. Newer rivals write smarter summaries; Otter still feels like transcription with AI layered on top.","limitations":["Accuracy dips with crosstalk and accents.","Bot-in-meeting creates friction with external guests.","AI summaries are generic."],"strengths":["Solid transcription and searchable meeting history.","Deep calendar integration.","Safe pick for teams already using it."],"use_for":["Teams that need transcription with calendar integration.","Meeting archives you need to search later."],"skip_when":["You have heavy crosstalk or accented speakers.","External guests resist a bot joining meetings.","You need smart, contextual summaries, not generic ones."],"receipts":["https://otter.ai","https://otter.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"parallel","name":"Parallel","tagline":"Web research APIs for agents — priced per request, not per token, with a free monthly cap.","url":"https://www.parallel.ai","category":"data","pricing":"freemium","price_note":"up to 5,000 requests/mo free; Search from $0.001–$0.005/10 results; Task API $0.005–$2.40/request; Extract $0.001/request","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Agent-web-search with a cost you can quote before the call. Search/Extract are cheap and sync; Task/Ultra processors get into dollars per request. Not Tavily and not Exa — those are other cards. SOC2 on the public table. Untested against a live agent loop here.","limitations":["Ultra8x Task is $2.40/request; \"deep research\" is not the $0.001 Search line.","Find All bills per match plus a processor fixed fee — easy to surprise.","5k free requests/mo is shared across APIs in the marketing table; confirm in console.","Not a general search engine UI; this is an API."],"strengths":["Search and Extract are cheap, synchronous APIs with predictable per-request pricing.","Task API covers deeper agent work when you can accept higher per-call cost.","SOC2 compliance is listed publicly."],"use_for":["Agent workflows that need structured web search or extraction.","Cost-sensitive projects using the 5k free monthly request cap."],"skip_when":["You expected deep research at the cheap Search price point.","Workflows use Find All where per-match billing can surprise.","You need a general-purpose search engine UI rather than an API."],"receipts":["https://www.parallel.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"pen-dev","name":"pen.dev","tagline":"Canvas-based design tool that exports clean production code.","url":"https://pen.dev","category":"design","pricing":"free","price_note":"currently free; paid plans may be added later","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"The canvas-to-code pitch is appealing for quick mocks, but the generated code still needs engineering review. Best for prototypes and early design exploration.","limitations":["Generated code rarely lands in production without cleanup.","Smaller community than established design tools.","Collaboration features lag behind Figma."],"strengths":["Canvas-based design tool that exports clean production-oriented code.","Quick mocks and early design exploration without leaving the canvas."],"use_for":["Prototypes and early design exploration before engineering handoff.","Designers who want to generate starter code from a canvas layout."],"skip_when":["You expect generated code to ship to production without cleanup.","Team collaboration workflows depend on Figma-level features.","You need a large community and plugin ecosystem."],"receipts":["https://pen.dev","https://pen.dev/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"penpot","name":"Penpot","tagline":"Open-source design and prototyping platform built for designers and developers.","url":"https://penpot.app","category":"design","pricing":"freemium","price_note":"Professional free for up to 8 editors/10GB; Unlimited $7/user/mo; Enterprise from $950/mo; Private Server $50k/yr","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"A genuine open-source Figma alternative with CSS-friendly layout tools and dev handoff. The free plan covers small teams generously, and you can self-host. Plugin ecosystem and advanced prototyping still trail Figma.","limitations":["Plugin and template ecosystem is smaller than Figma's.","Advanced prototyping and animation features are more limited.","Self-hosting requires infrastructure; the managed free plan has storage caps."],"strengths":["Genuine open-source Figma alternative with CSS-friendly layout tools.","Dev handoff is built in, not an afterthought.","Free plan covers small teams generously.","Self-host option keeps design data in-house."],"use_for":["Teams wanting an open-source design and prototyping tool.","Small teams that fit the free plan limits."],"skip_when":["You need Figma's plugin and template ecosystem.","Advanced prototyping and animation are required.","You cannot operate a self-hosted instance."],"receipts":["https://penpot.app","https://penpot.app/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"perplexity","name":"Perplexity","tagline":"Answer engine with citations — search rebuilt around LLMs.","url":"https://www.perplexity.ai","category":"data","pricing":"freemium","price_note":"free tier; Pro $20/mo, Max $200/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["gpt-5-6-terra","claude-sonnet-5","gemini-3-1-pro-preview"],"verdict":"ship","verdict_text":"Replaced Google for a real share of day-to-day queries: cited answers, follow-up context, and model choice on Pro. The receipt-link habit fits the catalog mindset.","limitations":["Citations occasionally point at sources that do not support the claim.","Complex queries eat through Pro search limits.","Publisher disputes raise long-term sustainability questions.","Comet browser went free for all (2026-03)."],"strengths":["Cited answers that replaced Google for a real share of day-to-day queries.","Follow-ups and Pro model choice fit a research habit, not a chatbot toy."],"use_for":["Everyday questions that should come back with sources.","A first research pass before you open the primary docs."],"skip_when":["Citations that do not support the claim would be dangerous.","Complex queries would eat Pro search limits.","Publisher-dispute risk is a blocker for how you rely on it."],"receipts":["https://www.perplexity.ai","https://docs.perplexity.ai/guides/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"pgvector","name":"pgvector","tagline":"Open-source vector similarity search extension for Postgres.","url":"https://github.com/pgvector/pgvector","category":"data","pricing":"free","free_tier":true,"open_source":true,"api":false,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Requires a running Postgres server and manual extension build/install; hosted provider support varies."],"receipts":["https://github.com/pgvector/pgvector"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"pharos","name":"Pharos","tagline":"Package manager and registry for MCP servers, with dependency resolution and a local daemon.","url":"https://getpharos.dev/","category":"dev-infra","pricing":"free","price_note":"MIT open source CLI; registry is free to use","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"situational","verdict_text":"Brings npm-like semantics to MCP servers: search across multiple registries, install with dependency resolution, lock versions, and run a local daemon that JIT-loads HTTP/SSE servers. The safe config writes and broad client support make it the most credible attempt yet to tame the growing MCP-server zoo.","limitations":["Registry and discovery are new; package quality varies.","stdio servers are not managed by the daemon; clients still handle those as child processes.","Publishing requires GitHub OAuth and acceptance of a young registry's policies."],"strengths":["npm-like package management for MCP servers with dependency resolution and version locking.","Local daemon JIT-loads HTTP/SSE MCP servers and keeps client configs clean.","Broad client support makes it practical across different MCP consumers."],"use_for":["Teams managing many MCP servers and wanting install/lock semantics.","Setups that rely on HTTP/SSE MCP servers served locally."],"skip_when":["You mostly use stdio MCP servers the daemon cannot manage.","You are not comfortable trusting a young registry's policies.","You cannot accept OAuth-based publishing requirements."],"receipts":["https://getpharos.dev/","https://github.com/Wpnx330/pharos-cli"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"photoroom","name":"Photoroom","tagline":"Product-photo editor — background removal plus credit-metered generative tools for listings.","url":"https://www.photoroom.com","category":"image","pricing":"freemium","price_note":"Free Space (no commercial use); paid yearly Pro 8,000 AI credits + 1,000 exports/mo, Max 25,000 + 3,000, Ultra 75,000 + 10,000; USD is JS-rendered on the pricing page — confirm in-browser; API/MCP billed separately","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"The listing-photo hammer: core retouch/background tools are in the paid plans; Virtual Model / video / Edit-with-AI burn a shared credit pool. This fetch saw credits and export caps on photoroom.com/pricing, not a stable HTML dollar figure, so the card will not copy blog \"$7.50/mo\" numbers. Free is not a commercial license. API is a second wallet.","limitations":["Dollar amounts are JS-rendered; do not budget from third-party blogs.","Free Space forbids commercial use; seats follow the Space, not the user.","Credit-heavy tools (video, fashion models, Edit with AI) exhaust the pool fast.","REST/MCP usage is not the app export quota — it's a separate credit wallet."],"strengths":["Purpose-built for product listing photos and background removal.","Paid plans bundle core retouch tools with export quotas."],"use_for":["E-commerce sellers polishing product photos for listings.","Teams that want a visual editor rather than a generative-only tool."],"skip_when":["You need stable, scrapable USD pricing for budget approval.","Free-tier commercial use is required — the Free Space forbids it.","Heavy use of video, virtual models or Edit-with-AI would exhaust the credit pool."],"receipts":["https://www.photoroom.com/pricing","https://help.photoroom.com/en/articles/6976012-what-are-photoroom-s-plans"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"pi","name":"Pi","tagline":"Minimal open-source coding-agent harness — TUI, SDK, RPC, BYOK, customize instead of feature-bloat.","url":"https://pi.dev","category":"coding","pricing":"free","price_note":"CLI/SDK free; you pay the model provider; packages install from npm or git","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"Mario Zechner's harness: small prompt, tree-structured sessions, 15+ providers, and a bias toward \"build the feature as an extension\" instead of shipping MCP/subagents/plan mode. Powerful if you like assembling; empty if you wanted Cursor. Not a Ship without using it on a repo.","limitations":["No bundled MCP, sub-agents, or permission UI — by design, you add them.","Quality is 100% the model you authenticate.","Docs and packages assume comfort with git/npm and a terminal.","Not Inflection Pi and not π the number — this is pi.dev / pi-mono."],"strengths":["Minimal open-source harness with TUI, SDK, and RPC in one repo.","Supports 15+ providers and is built for extension over feature-bloat.","Free CLI/SDK; you only pay the model provider you authenticate."],"use_for":["Developers who want to assemble and customize their own coding agent.","Terminal-first workflows comfortable with git/npm and BYOK setup."],"skip_when":["You want bundled MCP, sub-agents, or permission UI out of the box.","You need polished GUI docs and a guided onboarding experience.","You are not willing to bring and manage your own model keys."],"receipts":["https://pi.dev","https://github.com/badlogic/pi-mono"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"pika","name":"Pika","tagline":"Fast, playful video generation with viral creative effects (Pikaffects).","url":"https://pika.art","category":"video","pricing":"freemium","price_note":"Basic free (80 credits/mo); Standard $8/mo, Pro $28/mo, Fancy $76/mo billed yearly; commercial use starts at Pro","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Best for social-native effects and quick meme or short-form content. The Pikaffects gimmicks work for engagement. For cinematic or realistic footage it lags Runway, Kling and Veo, and lower tiers carry watermarks and resolution limits.","limitations":["Realism below Runway, Kling and Veo for live-action-style footage.","Short clip lengths per generation.","Watermark and resolution limits on lower tiers."],"strengths":["Fast video generation with viral creative effects (Pikaffects).","Strong for social-native meme and short-form content.","Cheap entry point with $8/mo Standard tier."],"use_for":["Quick social clips and meme content that leans on effects.","Experiments where engagement matters more than realism."],"skip_when":["You need cinematic or realistic live-action footage.","Watermarks and resolution limits on lower tiers are unacceptable.","Clip length per generation is too short for the story."],"receipts":["https://pika.art","https://future-stack-reviews.com/pika-labs-review/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"pinecone","name":"Pinecone","tagline":"Managed vector database — the serverless default for embeddings at scale.","url":"https://www.pinecone.io","category":"data","pricing":"freemium","price_note":"free Starter; Builder $20/mo; Standard pay-as-you-go (min $50/mo)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"The smoothest managed vector DB to operate, and the free tier covers prototypes. But self-hosted Qdrant or pgvector now handle most workloads for a fraction of the cost — pay only if ops time is your bottleneck.","limitations":["Costs scale poorly versus self-hosted alternatives.","Proprietary; migrating away is a rewrite.","Pricing-model changes have burned users before."],"strengths":["Smoothest managed vector database to operate at scale.","Free Starter tier covers most prototypes without upfront cost.","Serverless architecture removes cluster provisioning overhead."],"use_for":["Prototypes and production apps where ops time is the bottleneck.","Teams that would rather not run Qdrant or pgvector themselves."],"skip_when":["Cost matters more than the managed-vector-DB convenience.","You may need to migrate off a proprietary API later.","Past pricing changes have already burned your budget."],"receipts":["https://www.pinecone.io","https://www.pinecone.io/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"pipedream","name":"Pipedream","tagline":"Workflow automation plus Connect APIs for embedding 3,000+ apps into your own product.","url":"https://pipedream.com","category":"automation","pricing":"freemium","price_note":"Free plan with daily credit cap; paid platform fee includes a credit base — see pipedream.com/pricing; Connect also bills unique end users","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Zapier for people who will write Node in a step. Credits are 30s of 256MB compute, not per-step. Connect is a second product: credits plus unique end users. Docs are the receipt; the marketing pricing page is JS-heavy. Prototype on Free, don't pretend the daily cap is production.","limitations":["Free has a daily credit ceiling; overage requires a paid plan.","Memory bumps multiply credits (512MB = 2×).","Connect production and unique-user billing are easy to miss at prototype time.","Custom private event sources are not covered by the \"source run is free\" rule."],"strengths":["Write Node.js in a step for code-level flexibility.","3,000+ app integrations.","Connect APIs for embedding automations in your product."],"use_for":["Developers who want scriptable workflow automation.","Products that need embedded integrations via Connect."],"skip_when":["You need simple no-code automation without daily limits.","The free daily credit cap is too low for your volume.","Connect's unique end-user billing would surprise the budget."],"receipts":["https://pipedream.com/docs/pricing","https://pipedream.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"playwright-mcp","name":"Playwright MCP","tagline":"Microsoft's MCP server for browser automation via accessibility snapshots — no vision model required.","url":"https://github.com/microsoft/playwright-mcp","category":"mcp","pricing":"free","price_note":"Apache-2.0; npx @playwright/mcp@latest; you pay the model, not Microsoft","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"ship","verdict_text":"The default browser MCP: structured accessibility snapshots instead of screenshots, npx install, published to the official MCP registry on each release. Still a 0.0.x line, and it launches a real browser on your machine.","limitations":["Version line is still 0.0.x (v0.0.79 as of 2026-08-06).","Launches a real browser on the host — stdio is a local-privilege process.","Not a replacement for Playwright Test; this is agent control, not CI."],"strengths":["Official Microsoft server with a11y snapshots — no vision model required.","One-line install: npx @playwright/mcp@latest; works with Cursor, Claude, VS Code.","Released to the official MCP registry on each tag."],"use_for":["Agents that need to click through a live page without screenshot tokens.","Coding agents that already speak MCP and need a browser, not a test runner."],"skip_when":["You need Chrome DevTools traces, heap snapshots, or a live debugging session.","You cannot let an agent drive a local browser process.","You wanted the unmaintained ExecuteAutomation fork instead of Microsoft's."],"receipts":["https://github.com/microsoft/playwright-mcp","https://playwright.dev/docs/getting-started-mcp"],"affiliate":"none","evidence_tier":"source-verified","guide":{"install":"npx @playwright/mcp@latest — then attach the stdio server in the harness MCP config.","requirements":["A desktop/CI that can launch Chromium","A harness that speaks MCP"],"cost":"The server is Apache-2.0. You pay the model tokens and the machine that runs the browser.","security":"This launches a real browser on the host. Stdio MCP is a local-privilege process. Do not attach it to an agent that talks to strangers.","breaks_when":["You needed DevTools traces or heap snapshots — use chrome-devtools-mcp instead.","The environment cannot launch a browser (some CI images)."]},"momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"postgres-mcp","name":"Postgres MCP","tagline":"Reference MCP server for querying Postgres — useful and easy to point at production by mistake.","url":"https://github.com/modelcontextprotocol/servers","category":"mcp","pricing":"free","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: the official servers monorepo moves; pin a path/commit.","Read-write SQL from an agent is a high-privilege boundary. Use a read replica or skip."],"receipts":["https://github.com/modelcontextprotocol/servers"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"promptfoo","name":"Promptfoo","tagline":"Open-source CLI and library for evaluating, benchmarking, and red-teaming LLM apps.","url":"https://www.promptfoo.dev","category":"dev-infra","pricing":"freemium","price_note":"Community plan is free forever with all core evaluation features and 10,000 red-team probes per month; Enterprise and On-Premise are custom.","free_tier":true,"open_source":true,"api":false,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Community red teaming is capped at 10,000 probes per month and omits team sharing, continuous monitoring, and managed cloud."],"receipts":["https://www.promptfoo.dev","https://www.promptfoo.dev/docs/","https://www.promptfoo.dev/pricing/"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"pydantic-ai","name":"PydanticAI","tagline":"Pydantic's typed Python agent SDK — models, tools, evals, OpenTelemetry, not a hosted studio.","url":"https://ai.pydantic.dev","category":"agents","pricing":"free","price_note":"MIT SDK free; you pay model/provider bills; Pydantic Logfire observability is a separate product","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"How Python people actually want to write agents: Pydantic models in, Pydantic models out, graph/durable exec if you need it. Tiny compared with LangGraph's sprawl. You still own the loop, keys, and evals. Not a replacement for a hosted agent platform.","limitations":["No bundled hosting; production is your process + your tracing backend.","Logfire is optional and billed separately if you use it.","Ecosystem is younger than LangChain; fewer copy-paste recipes.","Type-safety doesn't stop a model from calling the wrong tool."],"strengths":["Pydantic-native types in and out, matching how Python teams already validate data.","Graph and durable execution options without LangGraph's ecosystem sprawl.","You keep control of the loop, model keys, and evals."],"use_for":["Python teams building typed agents where validation is non-negotiable.","Projects that need self-hosted execution rather than a managed platform."],"skip_when":["You want bundled hosting and a managed agent studio.","You need LangChain's larger recipe ecosystem today.","You expect type safety to prevent all model tool-calling mistakes."],"receipts":["https://ai.pydantic.dev","https://github.com/pydantic/pydantic-ai"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"qdrant","name":"Qdrant","tagline":"Rust-built open-source vector DB — fast, lean and easy to self-host.","url":"https://qdrant.tech","category":"data","pricing":"freemium","price_note":"self-host free; managed cloud free tier, paid usage-based","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"ship","verdict_text":"The pragmatic vector DB pick: single binary, low footprint, predictable performance, and a free cloud tier that is actually usable. Most RAG prototypes should start here.","limitations":["Fewer built-in retrieval features than Weaviate; bring your own reranking.","Clustering and sharding maturity behind the incumbents.","Smaller ecosystem and fewer integrations than Pinecone."],"strengths":["Pragmatic vector DB: single binary, low footprint, predictable performance.","Free cloud tier is actually usable; most RAG prototypes should start here."],"use_for":["A first RAG store you can self-host or try managed.","Teams that want a lean Rust engine more than a feature zoo."],"skip_when":["You need Weaviate-style built-in retrieval and reranking.","Clustering and sharding have to match incumbent maturity.","The team needs Pinecone's wider integrations."],"receipts":["https://github.com/qdrant/qdrant"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"quillbot","name":"QuillBot","tagline":"Paraphrase / grammar / citation helper, still a student-tool more than an agent stack.","url":"https://quillbot.com","category":"writing","pricing":"freemium","free_tier":true,"open_source":false,"api":false,"self_host":false,"limitations":["Radar: paraphrasers are a crowded, low-trust category.","Citation features need a primary-source check, not this card."],"receipts":["https://quillbot.com"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"qwen-api","name":"Qwen API","tagline":"Alibaba Cloud Model Studio API for text, code, image, audio, video and omni models with OpenAI compatibility.","url":"https://www.alibabacloud.com/help/en/model-studio/what-is-model-studio","category":"models-api","pricing":"paid","price_note":"Qwen3.7 Max ~$1.48/$4.43 per 1M, Qwen3.6 Flash $0.19/$1.13 per 1M, Qwen-Image 2.0 ~$0.04/image; Token Plan and Coding Plan subscriptions available","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["qwen3-8-max"],"model_routing":"locked","verdict":"ship","verdict_text":"The widest Chinese model menu in one API, with open weights and a real international region. Pricing is attractive, but the console still feels built for China-first users, so expect some billing and rate-card friction.","limitations":["The hosted Model Studio API is proprietary; many Qwen weights are open releases available separately.","International vs China pricing diverges; OpenRouter can be cheaper for some models.","Large MoE models are expensive to self-host.","China-region compliance and billing can be hard for non-China users.","Some docs and rate-card pages are JS-rendered or region-locked."],"strengths":["Widest Chinese model menu in one API, with open weights and an international region.","Pricing is attractive next to US frontier labs."],"use_for":["Qwen-family text, code and omni calls with OpenAI-compatible endpoints.","Workloads that can live on Alibaba Cloud's international region."],"skip_when":["China-region billing or compliance is the only path you get.","OpenRouter is already cheaper for the models you need.","You needed OSI-open hosted inference, not a proprietary studio API."],"receipts":["https://www.alibabacloud.com/help/en/model-studio/what-is-model-studio","https://www.alibabacloud.com/help/en/model-studio/model-pricing","https://llmcostcalc.com/cloud/qwen","https://a8gent.com/models/alibaba-qwen"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"qwen-code","name":"Qwen Code","tagline":"Open-source terminal coding agent from Alibaba, forked from Gemini CLI.","url":"https://github.com/QwenLM/qwen-code","category":"coding","pricing":"freemium","price_note":"CLI is free/open source; hosted free tier restricted to ~100 requests/day as of April 2026; Coding Plan Pro ~$50/month (~90K requests/month); BYOK available","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"The CLI is free and open source, so it costs nothing to try with your own keys. The hosted free tier is basically gone, and the paid coding plan is mid-priced. Best for teams that already use Qwen or want a Gemini-CLI-like workflow.","limitations":["Free hosted inference was cut to roughly 100 requests/day in April 2026.","Less mature agent ecosystem than Claude Code or Kimi Code.","Configuration surface is larger than single-vendor tools.","Strongest when paired with Qwen models; other providers are supported but less optimized."],"strengths":["CLI is free and open source — costs nothing to try with your own keys.","Gemini-CLI-shaped workflow if that is already how you work.","Strongest on Qwen models; BYOK keeps the hosted plan optional."],"use_for":["Teams already on Qwen who want a terminal agent, not another IDE.","Self-hosted / BYOK coding where a $50 plan is the fallback, not the start."],"skip_when":["You expected a generous hosted free tier — that was cut in April 2026.","You need the mature agent ecosystem around Claude Code or Kimi Code.","You do not want to operate a larger config surface."],"receipts":["https://github.com/QwenLM/qwen-code","https://aicoolies.com/reviews/qwen-code-review","https://aiidelist.com/ide/qwen-code","https://inventivehq.com/blog/qwen-code-still-free-2026-shutdown"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"qwen-image","name":"Qwen-Image","tagline":"Alibaba Cloud text-to-image and image-editing models, including Qwen-Image 2.0 and Plus/Max variants.","url":"https://github.com/QwenLM/Qwen-Image","category":"image","pricing":"paid","price_note":"Qwen-Image 2.0 ~$0.029/image, Image Plus ~$0.0287/image, Image Edit Max similar per-image rate","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"Cheap image generation and editing through the same Alibaba Cloud account as Qwen text. It's not Midjourney or top-tier Flux, but the price and text-in-image accuracy make it a practical choice for apps and bulk workflows.","limitations":["Quality tier below top closed image models.","Western awareness and community resources are lower than Midjourney/Flux.","China-region compliance and data-residency rules apply.","Editing endpoints are priced separately and can add up."],"strengths":["Cheap per-image generation and editing through an existing Alibaba Cloud account.","Text-in-image accuracy is a practical edge over some cheaper rivals.","Useful for apps and bulk workflows that prioritize cost over absolute quality."],"use_for":["Bulk image generation or editing pipelines that need low unit cost.","Apps already using Qwen text models and Alibaba Cloud services."],"skip_when":["You need top-tier image quality comparable to Midjourney or best Flux.","China-region compliance and data-residency rules are a hard no.","Editing volume would make the separate editing endpoint costs add up."],"receipts":["https://github.com/QwenLM/Qwen-Image","https://gate.ai/blog/qwen-image-2-0-qwen-specs-pricing-api-use-cases","https://gate.ai/blog/qwen-image-plus-specs-pricing-api-use-cases-2026","https://developer.puter.com/tutorials/qwen-api-pricing/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"radix-ui","name":"Radix UI","tagline":"Low-level, accessible UI primitives for React.","url":"https://radix-ui.com/primitives","category":"design","pricing":"free","price_note":"MIT open source","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"ship","verdict_text":"The foundation most modern React component libraries are built on. Primitives are accessible, unstyled, and behave correctly across browsers.","limitations":["You must style everything yourself.","Not a drop-in component library for non-technical users.","Some primitives have a learning curve."],"strengths":["Accessible, unstyled React primitives that behave across browsers.","The foundation most modern React kits — including shadcn — sit on."],"use_for":["Building a design system on correct behavior, then styling it yourself.","Teams that want a11y baked into the primitive, not bolted on."],"skip_when":["You need a drop-in styled library for non-engineers.","Styling everything from scratch is not in the budget.","The primitive API learning curve would stall the sprint."],"receipts":["https://radix-ui.com/primitives","https://github.com/radix-ui/primitives"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"ragflow","name":"RagFlow","tagline":"Open-source RAG engine that ingests multi-format data and combines hybrid search with agent orchestration.","url":"https://ragflow.io","category":"data","pricing":"freemium","price_note":"Free tier: 5 apps, 1 team member, 0.1 GB storage, 500 credits/month; paid tiers from $29/month.","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","API key access is unavailable on the free tier; some deployment options require paid plans."],"receipts":["https://ragflow.io"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"raycast","name":"Raycast","tagline":"macOS launcher with built-in AI chat, commands and a huge extension store.","url":"https://www.raycast.com","category":"productivity","pricing":"freemium","price_note":"free tier (50 AI messages incl.); Pro $10/mo ($8/mo yearly)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"ship","verdict_text":"If you are on macOS, this is hard to beat: AI, clipboard history, snippets and window management in one keystroke. The free tier is generous; the AI quality depends on what you pay for.","limitations":["macOS + Windows only; no Linux build.","AI model quality depends on your plan tier.","Extension quality varies wildly."],"strengths":["macOS launcher that actually replaces a pile of utilities: AI, clipboard, snippets, windows.","Free tier is generous; one keystroke is the product."],"use_for":["Daily driver launcher on a Mac.","Light AI chat from the keyboard without opening a browser."],"skip_when":["You need Linux support.","AI quality on your plan tier is not good enough.","You would depend on extensions of wildly uneven quality."],"receipts":["https://www.raycast.com","https://www.raycast.com/pricing","https://www.raycast.com/changelog/windows","https://manual.raycast.com/ai/usage-limits"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"react-bits","name":"React Bits","tagline":"Copy-paste animated React components for landing pages and interactions.","url":"https://reactbits.dev","category":"design","pricing":"free","price_note":"MIT + Commons Clause; source available, free for personal and commercial use","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"ship","verdict_text":"A solid source of polished animated components you can copy straight into a React project. The motion quality is consistently good, but the license is MIT plus a Commons Clause, so check the terms before redistributing.","limitations":["Tailwind-first styling assumes your project is already on that stack.","Some components need accessibility tweaks out of the box.","No organized design system; pick-and-choose only."],"strengths":["Polished animated React components you can paste into a landing page today.","Motion quality is consistently good for the copy-paste workflow."],"use_for":["Landing-page interactions on an existing Tailwind React app.","Pick-and-choose motion without adopting a full design system."],"skip_when":["Commons Clause would block how you redistribute.","The project is not on Tailwind.","Components need to be accessible without a manual pass."],"receipts":["https://reactbits.dev","https://github.com/DavidHDev/react-bits"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"reclaim","name":"Reclaim.ai","tagline":"AI calendar that defends focus time, habits and tasks automatically.","url":"https://reclaim.ai","category":"productivity","pricing":"freemium","price_note":"free tier; from $10/user/mo (annual)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"ship","verdict_text":"Focus time, habits and tasks slot themselves around your meetings, and the free tier covers a lot. It works with both Google and Outlook calendars, though the richest integrations still lean Google.","limitations":["Calendar-centric by design; deepest automations still lean Google-side.","Can over-fill your calendar if misconfigured.","Team features need whole-team buy-in."],"strengths":["Defends focus time, habits and tasks around the meetings you already have.","Free tier covers a lot; works with Google and Outlook."],"use_for":["Solo or small-team calendars that keep getting eaten by meetings.","Habits and tasks that should auto-schedule, not live in a separate app."],"skip_when":["The deepest automations you need are Google-only and you are not on Google.","Misconfiguration would over-fill the calendar.","Team features need buy-in you do not have."],"receipts":["https://reclaim.ai","https://reclaim.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"recraft","name":"Recraft","tagline":"Design-focused image model with vector output and brand style controls.","url":"https://www.recraft.ai","category":"image","pricing":"freemium","price_note":"free tier; from $10/mo (Basic)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Pick it for product and marketing design: vector SVG output, style consistency and brand palettes are things Midjourney does not handle. Skip it for photorealism or open-ended art; the credit math is also annoying to budget.","limitations":["Photorealism trails the category leaders.","Credit system is confusing to budget.","Smaller community and fine-tune ecosystem."],"strengths":["Vector SVG output for product and marketing design.","Style consistency and brand palette controls.","Free tier exists; paid starts at $10/mo."],"use_for":["Product design and marketing assets that need editable vector output.","Brand work where palette and style consistency matter more than photorealism."],"skip_when":["Photorealism or open-ended art is the main output.","You need predictable credit budgeting; the system is confusing.","You want a large fine-tune or community ecosystem."],"receipts":["https://www.recraft.ai","https://www.recraft.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"refero-styles","name":"Refero Styles","tagline":"Design example library and prompts for steering AI agents toward better UI.","url":"https://styles.refero.design","category":"design","pricing":"freemium","price_note":"free browse; paid plans unlock full library","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Handy when you need concrete examples to show an AI what \"good\" looks like. The curation is decent, but it is still a reference gallery with prompts — not components you drop into a repo.","limitations":["Most useful examples sit behind a paywall.","Not a build tool; you still translate ideas into your own stack.","Coverage is uneven across categories."],"strengths":["Curated design examples that help steer AI agents toward better UI.","Prompts and reference gallery for showing what 'good' looks like.","Useful when you need concrete references before writing components."],"use_for":["Teams that want visual references before building components.","Early design exploration where examples matter more than a component library."],"skip_when":["You need drop-in components that ship straight into a repo.","The category you care about has thin coverage.","The most useful examples sit behind the paywall."],"receipts":["https://styles.refero.design","https://refero.design/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"reflect","name":"Reflect","tagline":"AI-native note-taking app with networked thoughts and end-to-end encryption.","url":"https://reflect.app","category":"productivity","pricing":"paid","price_note":"from $10/mo","free_tier":false,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Fast, keyboard-driven notes with AI summarization and strong privacy. The daily notes workflow is excellent for writers and thinkers. It is smaller and pricier than Notion, with fewer integrations.","limitations":["No free tier.","Integration ecosystem is small.","Mobile app is less mature than desktop."],"strengths":["Fast, keyboard-driven daily notes workflow.","Built-in AI summarization.","End-to-end encryption for your notes."],"use_for":["Writers and thinkers who want a daily notes habit.","Users who prioritize privacy in note-taking."],"skip_when":["You need a free tier beyond the 14-day trial.","Rich third-party integrations are essential.","Your workflow is mobile-first."],"receipts":["https://reflect.app","https://reflect.app/privacy"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"relay-app","name":"Relay.app","tagline":"AI-native workflow product with human-in-the-loop steps, aimed at ops not iPaaS veterans.","url":"https://www.relay.app","category":"automation","pricing":"freemium","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: vendor marketing is ahead of our reproduction.","Human-in-the-loop is the feature — it will not fully replace n8n."],"receipts":["https://www.relay.app"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"relevance-ai","name":"Relevance AI","tagline":"No-code platform for building AI agent workforces for ops teams.","url":"https://relevanceai.com","category":"agents","pricing":"freemium","price_note":"free tier; paid plans usage-based from ~$19/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Useful when ops or sales needs research and outreach agents without calling a developer. Engineers hit the abstraction ceiling fast and will want something like CrewAI or LangGraph. Pricing also jumps quickly once the free tier is gone.","limitations":["Pricing jumps steeply past the free tier.","Black-box agent runs are hard to debug.","Workflows built on it are locked into the vendor."],"strengths":["No-code platform for building research and outreach agent workforces.","Useful for ops and sales teams that need agents without calling a developer.","Bundled API access for running agent tools and workflows."],"use_for":["Ops or sales teams automating research and outreach without engineering help.","Non-technical builders who want a visual agent workforce platform."],"skip_when":["Engineers need transparent, debuggable agent runs beyond the abstraction ceiling.","Pricing past the free tier jumps quickly for ongoing production usage.","Vendor lock-in is unacceptable for long-lived workflows."],"receipts":["https://relevanceai.com","https://relevanceai.com/docs/get-started/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"relume","name":"Relume","tagline":"Webflow site builder and component library for marketing websites.","url":"https://relume.io","category":"design","pricing":"freemium","price_note":"free library browse; paid plans for site builder","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Strong for agencies building Webflow sites at speed. The component library is large, but the value drops if you are not in the Webflow ecosystem.","limitations":["Best inside Webflow; export to other stacks is limited.","Paid plans are required for real site building.","Designs can look samey without custom work."],"strengths":["Large component library purpose-built for Webflow marketing sites.","Free library browse lets you inspect components before paying.","Speeds up agency builds when the stack is already Webflow."],"use_for":["Agencies building Webflow marketing sites at speed.","Teams already committed to the Webflow ecosystem."],"skip_when":["You are not using Webflow; export to other stacks is limited.","You need real site-building features without a paid plan.","You cannot invest in custom work to avoid samey templates."],"receipts":["https://relume.io","https://relume.io/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"replicate","name":"Replicate","tagline":"Run any community model via API — pay per second, no GPU wrangling.","url":"https://replicate.com","category":"dev-infra","pricing":"paid","price_note":"pay per run; image models from ~$0.003/image","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"ship","verdict_text":"A quick way to run models you don't want to host yourself — thousands of community models behind one API, billed per second. Good for prototypes and workloads that spike. Steady-state use gets expensive, and quality varies because it's community-published.","limitations":["Per-second pricing gets expensive at steady volume.","Cold starts on less-popular models.","Model quality and maintenance vary by community author.","Acquired by Cloudflare (2025-11); API and pricing unchanged so far."],"strengths":["Thousands of community models behind one API, billed per second, no GPU wrangling.","Good for prototypes and spiky workloads you do not want to host."],"use_for":["Trying a community model this afternoon.","Burst inference that should die when the job ends."],"skip_when":["Steady volume would make per-second pricing the expensive path.","Cold starts on obscure models would show up to users.","You need guaranteed maintenance, not a community author."],"receipts":["https://replicate.com","https://replicate.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"replit-agent","name":"Replit Agent","tagline":"Replit's browser-based AI app builder with built-in deployment and effort-based credits.","url":"https://replit.com/agent","category":"coding","pricing":"freemium","price_note":"Starter free with limited daily credits; Core $25/mo ($20/mo annual) with usage credits; Pro $100/mo; Enterprise custom","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Fastest zero-setup path from prompt to deployed app, now with Agent 4 and a design canvas. The effort-based credit model is fairer than flat checkpoints, but costs scale with task complexity. Browser IDE still lags behind local tools for serious engineering.","limitations":["Agent tasks burn credits that scale with complexity; bills can surprise heavy users.","Browser editor and debugger are less capable than VS Code or JetBrains.","Production hosting is mostly Replit's platform; exporting to external infra takes work."],"strengths":["Fastest zero-setup path from prompt to a deployed app in a browser IDE.","Agent 4 and a design canvas speed up end-to-end app building.","Effort-based credits are fairer than flat checkpoint pricing."],"use_for":["Beginners and prototypers who want deployment without local setup.","Small apps and MVPs where a browser editor is good enough."],"skip_when":["Serious engineering needs the debugger and extensions of VS Code or JetBrains.","You expect to export and run production workloads outside Replit's hosting.","Agent tasks with heavy complexity would burn through credit allowances."],"receipts":["https://replit.com/agent","https://replit.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"resemble-ai","name":"Resemble AI","tagline":"Deepfake detection and synthetic-media security for audio, image, video and meetings.","url":"https://www.resemble.ai","category":"audio","pricing":"freemium","price_note":"free Flex tier (pay-as-you-go); Team from $350/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Strong detection and watermarking stack for teams worried about synthetic-media fraud. The Flex tier is genuinely free to start, but production use gets expensive fast. Voice generation is no longer the headline; verify it still fits your pipeline.","limitations":["Paid plans are expensive for casual use.","Voice and speech features are now secondary; check current capabilities.","Enterprise features require sales contact."],"strengths":["Strong detection and watermarking stack for synthetic-media fraud.","Flex tier is genuinely free to start for experimentation.","Covers audio, image, video and meeting deepfake scenarios."],"use_for":["Teams worried about synthetic-media fraud across audio, image and video.","Security workflows that need watermarking and detection tooling."],"skip_when":["You are looking for a casual, low-cost production tool.","Voice generation is still your primary need.","You need enterprise features without a sales contact."],"receipts":["https://www.resemble.ai","https://www.resemble.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"rev-ai","name":"Rev AI","tagline":"Developer-first speech-to-text API offering accurate transcription in 57+ languages at scale.","url":"https://rev.ai","category":"audio","pricing":"freemium","price_note":"Pay-as-you-go plan includes free credits equivalent to 5 hours of Reverb ASR usage; volume pricing is available.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Pay-as-you-go plan includes free credits equivalent to 5 hours of Reverb ASR usage."],"receipts":["https://rev.ai","https://www.rev.ai/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"reve","name":"Reve","tagline":"4K image model with layout-native edit/create — consumer energy plans plus a credit API.","url":"https://reve.art","category":"image","pricing":"freemium","price_note":"Free with daily energy; Lite $7.99/mo (5× energy); Pro $19.99/mo (100× energy, video energy); API from $10 = 7,500 credits","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Layout-aware image tool (create/edit/remix, annotations, live layers), not a Midjourney clone. Consumer math is energy, not seats: Lite $7.99, Pro $19.99 on Reve's help center. API is a second wallet in the console. Video energy exists on every plan with different limits (Free one-time allowance, Pro 250/mo). Untested on production brand work.","limitations":["\"Energy\" is opaque versus a per-image rate card; Lite/Pro multiply Free, they don't publish a raw count here.","API pricing lives in the console, not a public HTML table.","Training opt-out is Lite/Pro; Free is default opt-in.","Video is not a Pro-exclusive feature; energy limits differ by plan."],"strengths":["Layout-native image creation and editing with annotations and live layers.","Clear consumer plans at $7.99/$19.99 with published energy multiples.","A credit API for developers who want programmatic access."],"use_for":["Designers and creators who want layout-aware image editing, not prompt roulette.","Consumers comfortable with energy-based plans for image and video work.","Developers who need a credit API alongside the consumer editor."],"skip_when":["You need a transparent per-image rate card instead of \"energy\".","You want public API pricing outside the console.","You object to default opt-in model training on the free plan.","Video output matters and the energy limits don't fit your volume."],"receipts":["https://help.reve.com/hc/en-us/articles/46732445465108-Subscription-plans-overview","https://help.reve.com/hc/en-us/articles/46837930295316-Reve-API","https://help.reve.com/hc/en-us/articles/46731721501972-Model-training"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"roomote","name":"Roomote","tagline":"Source-available cloud coding teammate from the Roo Code team, working in chat and returning PRs with live previews.","url":"https://roomote.dev","category":"agents","pricing":"freemium","price_note":"Self-hosted free up to 10 users (Fair Core License, ALv2 after 2 years); paid per-capability tiers above; Roomote Cloud: 7-day trial, BYO inference key","free_tier":true,"open_source":false,"api":false,"self_host":true,"model_routing":"byok","limitations":["Radar: not field-run in this catalog wave.","Source-available Fair Core, not OSI open source; more than 10 users needs a paid license key.","Single-tenant design; cloud is bring-your-own-inference-key, so real cost is usage-based."],"receipts":["https://roomote.dev"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"runpod","name":"RunPod","tagline":"Serverless GPU cloud for deploying containerized inference endpoints that scale from zero to production.","url":"https://runpod.io","category":"models-api","pricing":"paid","price_note":"Serverless GPU workers from $0.58/hr to $9.98/hr, metered per second; pods and clusters billed hourly.","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"byok","limitations":["Radar: not field-run in this catalog wave.","You bring your own container and model; billing is per GPU-second/hour, not per token."],"receipts":["https://runpod.io","https://runpod.io/pricing","https://runpod.io/serverless"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"runway","name":"Runway","tagline":"The incumbent AI video suite — Gen-4.5 generation plus a full editing toolchain.","url":"https://runway.com","category":"video","pricing":"freemium","price_note":"free tier (125 one-off credits); Standard $12/mo, Pro $28/mo (yearly)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"Still the most complete AI video workflow. Gen-4.5 quality is competitive, and the editing tools — inpainting, motion brush, green screen — let you finish a shot instead of just generating a clip. Credit math gets painful if you iterate heavily, and the best quality modes sit behind pricier tiers.","limitations":["Credit math gets expensive for iteration-heavy projects.","Best quality modes sit behind pricier tiers.","Long coherent narratives still need heavy manual stitching.","Unlimited tier replaced by Max for new subscribers (2026-06)."],"strengths":["Most complete AI video workflow: generate plus inpainting, motion brush, green screen.","Gen-4.5 quality is competitive, and you can finish a shot in the same suite."],"use_for":["Short-form video that needs edit tools, not just a clip generator.","Teams that will pay to iterate inside one vendor."],"skip_when":["Credit math on heavy iteration would blow the budget.","Best quality modes sit behind a tier you will not buy.","You need long coherent narrative without manual stitching."],"receipts":["https://help.runwayml.com","https://runway.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":true,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"rytr","name":"Rytr","tagline":"Cheap consumer writing assistant with templates and a small free allowance.","url":"https://rytr.me","category":"writing","pricing":"freemium","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: vendor site only. Not compared in the field against Jasper/Sudowrite.","Template writers age badly when models jump."],"receipts":["https://rytr.me"],"affiliate":"none","evidence_tier":"radar","momentum":"redshift","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"sambanova","name":"SambaNova","tagline":"Enterprise AI platform and API for fast inference on open-source models.","url":"https://sambanova.ai","category":"models-api","pricing":"paid","price_note":"usage-based; enterprise and cloud tiers","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Fast inference with a focus on enterprise deployment and data sovereignty. Strong for regulated industries. The developer experience is less polished than API-first rivals, and self-serve onboarding is limited.","limitations":["Developer onboarding is more enterprise-focused.","Pricing is not as transparent as pure API providers.","Smaller community and example base than OpenAI/Anthropic."],"strengths":["Fast inference focused on enterprise deployment and data sovereignty.","Strong positioning for regulated industries.","Runs open-source models on dedicated hardware."],"use_for":["Regulated enterprises that need data-sovereign inference.","Teams prioritizing inference speed on open-source models.","Organizations with existing enterprise procurement channels."],"skip_when":["You need transparent, self-serve pricing and a polished developer UX.","You rely on a large community and example base.","You want to start with a free tier or small self-serve account."],"receipts":["https://sambanova.ai","https://docs.sambanova.ai"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"seedance","name":"Seedance","tagline":"ByteDance text/image-to-video model family served via Volcano Engine and third-party APIs.","url":"https://www.byteplus.com/en/product/Seedance","category":"video","pricing":"paid","price_note":"Seedance 2.0 ~$0.04–$0.056/s via third-party providers (5s 720p ≈ $0.05); Seedance 2.5 from $0.90/5s at 480p on some providers; no consumer subscription","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Seedance 2.0 is dirt-cheap for capable video, with good motion and prompt adherence. For Western users the path in is through resellers: ByteDance doesn't publish clean USD API pricing or a global self-serve console.","limitations":["Pricing is CNY-first; USD rates are reseller conversions.","Closed weights; no self-host fallback.","Western access and documentation are limited.","Seedance 2.5 pricing varies widely by provider and resolution."],"strengths":["Seedance 2.0 delivers capable video at very low cost per second.","Good motion quality and prompt adherence for the price.","Accessible via API for users who can navigate reseller channels."],"use_for":["Budget-conscious video generation where cost per clip matters most.","Teams that can source access through third-party API providers."],"skip_when":["You need transparent USD pricing from ByteDance directly.","Closed weights with no self-host fallback are a dealbreaker.","Western documentation and support channels are required."],"receipts":["https://dynalord.com/blog/seedance-pricing","https://www.atlascloud.ai/providers/bytedance","https://wavespeed.ai/seedance-2-5-api"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"seedream","name":"Seedream","tagline":"ByteDance's image generation model, served via BytePlus ModelArk and Dreamina.","url":"https://www.byteplus.com/en/product/Seedream","category":"image","pricing":"paid","price_note":"BytePlus ModelArk pay-as-you-go; Seedream 5.0 Pro packages from ~$6.99 intro; standard from $0.045/image","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Seedream 5.0 Pro produces strong prompt adherence and visual quality at aggressive per-image prices. Western access is through BytePlus ModelArk, with CNY-first billing and sparse USD documentation. A capable option for bulk asset generation if you accept ByteDance's trust boundary.","limitations":["Access and pricing are CNY-first; clean USD rate cards are hard to find.","Closed weights with no self-host or open-weights fallback.","Western signup, docs, and data-sovereignty concerns mirror other ByteDance AI products."],"strengths":["Seedream 5.0 Pro delivers strong prompt adherence and visual quality.","Aggressive per-image pricing, with packages starting around $0.045/image."],"use_for":["Bulk image asset generation where low cost per image matters.","Teams already comfortable with ByteDance's trust boundary for non-sensitive visuals."],"skip_when":["You need clean USD rate cards and Western billing documentation.","Closed weights with no self-host or open-weights fallback is a dealbreaker.","Data-sovereignty or Western signup concerns outweigh the price advantage."],"receipts":["https://www.byteplus.com/en/product/Seedream","https://ai.byteplus.com/product/Seedream"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"sentry-mcp","name":"Sentry MCP","tagline":"Sentry's remote MCP for coding agents — issues, traces, debug, not a full Sentry API.","url":"https://mcp.sentry.dev","category":"mcp","pricing":"free","price_note":"included with a Sentry account; hosted https://mcp.sentry.dev/mcp; stdio npx @sentry/mcp-server for self-hosted","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"Official Sentry remote MCP, aimed at human-in-the-loop coding agents, not a general Sentry bot. Hosted on Cloudflare. stdio exists for self-hosted Sentry and still needs a user auth token with write scopes.","limitations":["Tooling is for developer debug workflows, not full Sentry administration.","stdio AI search tools need OpenAI/Anthropic/OpenRouter or they disappear.","Self-hosted Seer and some skills are missing; disable them explicitly."],"strengths":["Official Sentry server at mcp.sentry.dev, built for Cursor / Claude Code.","Claude Code plugin with a delegated sentry-mcp subagent.","Remote by default; stdio for self-hosted Sentry."],"use_for":["Agents debugging an error that already lives in Sentry.","Teams that want issue/trace context inside the coding agent, not another tab."],"skip_when":["You needed the whole Sentry product API — this is a coding-agent slice.","AI search tools (search_events / search_issues) need a separate LLM provider.","You will not give event:write on a user auth token."],"receipts":["https://github.com/getsentry/sentry-mcp","https://mcp.sentry.dev"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":false},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"sequential-thinking-mcp","name":"Sequential Thinking MCP","tagline":"Official sequential-thinking MCP — a loop that spends tokens to look like reasoning.","url":"https://github.com/modelcontextprotocol/servers/tree/main/src/sequentialthinking","category":"mcp","pricing":"free","price_note":"MIT reference; npx -y @modelcontextprotocol/server-sequential-thinking","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"skip","verdict_text":"A stdio loop that asks the model to emit thought / nextThoughtNeeded / thoughtNumber. The model already thinks. This bills extra tokens for a ritual. Keep it off the default mcp.json.","limitations":["Each \"thought\" is another model round-trip you pay for.","Does not add tools, files, or a browser — only a structured monologue.","Easy to cargo-cult into every mcp.json because the name sounds serious."],"strengths":["Official reference, MIT, one npx line.","Visible thought steps in the host inspector if you are debugging a prompt."],"use_for":["Teaching someone what an MCP tool call looks like, not for production work.","A host inspector demo of repeated tool calls."],"skip_when":["You want better answers — spend the tokens on the task, not on a thinking costume.","The agent already plans (Claude, Codex, Cursor agent).","You are paying API rates and this would sit in every session."],"receipts":["https://github.com/modelcontextprotocol/servers/tree/main/src/sequentialthinking"],"affiliate":"none","evidence_tier":"source-verified","momentum":"redshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"shadcn-ui","name":"shadcn/ui","tagline":"Copy-paste React components built on Radix primitives and Tailwind CSS.","url":"https://ui.shadcn.com","category":"design","pricing":"free","price_note":"MIT open source","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"ship","verdict_text":"The closest thing to a standard React component stack right now. You own the code, theming is straightforward, and the ecosystem of blocks keeps growing.","limitations":["You maintain the components yourself after installation.","No packaged npm install; copy-paste only.","Quality of community blocks varies."],"strengths":["Closest thing to a standard React component stack: you own the code.","Theming is straightforward, and the block ecosystem keeps growing."],"use_for":["New React apps that should look consistent and stay in-repo.","Teams already on Tailwind and Radix."],"skip_when":["You wanted an npm package you do not have to maintain.","Copy-paste ownership is a non-starter for your org.","You would depend on uneven community blocks."],"receipts":["https://ui.shadcn.com","https://github.com/shadcn-ui/ui"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"shoehorn","name":"Shoehorn","tagline":"Quantize any language model to fit your machine's actual VRAM, then serve it locally.","url":"https://notactuallytreyanastasio.github.io/shoehorn/","category":"dev-infra","pricing":"free","price_note":"open source; bring your own compute and API keys","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"situational","verdict_text":"A practical answer to \"will this model run on my GPU?\" Shoehorn solves a per-tensor mixed-precision assignment from your real memory budget instead of forcing a preset quantization, squeezing out quality other presets waste. Free, local, and genuinely useful for anyone running LLMs on consumer hardware.","limitations":["Requires a working llama.cpp setup and some CLI comfort; the web UI still shells the same pipeline.","Windows support is compile-tested but not yet field-tested.","Auto-generated imatrices are skipped on Windows, so you need repos that publish one."],"strengths":["Quantizes any language model to fit your machine's actual VRAM budget.","Per-tensor mixed-precision assignment squeezes out quality other presets waste.","Free and local for running LLMs on consumer hardware."],"use_for":["Local LLM inference on consumer GPUs with tight VRAM constraints.","Scenarios where preset quantization throws away usable quality."],"skip_when":["You are not comfortable with llama.cpp and command-line workflows.","You run Windows and need auto-generated imatrices for every model.","A one-click web GUI is a hard requirement."],"receipts":["https://github.com/notactuallytreyanastasio/shoehorn"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"signoz","name":"SigNoz","tagline":"OpenTelemetry-native observability platform for traces, metrics, logs, and AI-agent telemetry.","url":"https://signoz.io","category":"dev-infra","pricing":"freemium","price_note":"Self-hosted Community edition is free; SigNoz Cloud Teams starts at $49 per month with $49 of usage credit; Enterprise starts at $4,000 per month.","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Cloud Teams has a $49 monthly minimum and bills per GB of traces/logs and per million metric samples beyond the included credit."],"receipts":["https://signoz.io","https://signoz.io/docs/","https://signoz.io/pricing/"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"singular","name":"Singular","tagline":"Autonomous multi-agent orchestration engine that runs coding agents in parallel against a repo.","url":"https://github.com/alex-reysa/singular-lite","category":"agents","pricing":"free","price_note":"GPL-3.0 open source; bring your own runner (Claude Code, Codex, etc.)","free_tier":true,"open_source":true,"api":false,"self_host":true,"verdict":"situational","verdict_text":"Brings git-worktree isolation, durable leases, and a three-tier scheduler to long-running agent workflows. It is engineered for teams that want many autonomous coding agents working on the same repo without stepping on each other. Setup is nontrivial and the documentation assumes you already run Claude Code or Codex.","limitations":["Requires bash 4+, Python, git, and a supported runner CLI already configured.","Engine must be installed from a checkout; there is no download mechanism for the pinned engine.","The default promoter is minimal, so real workflows need custom gate and promoter scripts."],"strengths":["Git-worktree isolation, durable leases and a three-tier scheduler for agent workflows.","Lets multiple autonomous coding agents work the same repo without collisions.","GPL-3.0 open source and self-hostable on your own runner."],"use_for":["Teams running multiple long-lived coding agents against one repository.","Shops already using Claude Code or Codex and wanting isolation and scheduling."],"skip_when":["You cannot install from source or configure a supported runner CLI.","Your workflows need a polished default promoter out of the box.","You lack the engineering time for nontrivial setup and custom gate scripts."],"receipts":["https://github.com/alex-reysa/singular-lite","https://github.com/alex-reysa/singular-lite/blob/main/LICENSE"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"snowflake-cortex","name":"Snowflake Cortex","tagline":"Managed LLM, RAG, and AI functions inside the Snowflake data cloud, priced in AI Credits.","url":"https://www.snowflake.com/en/data-cloud/cortex","category":"models-api","pricing":"paid","price_note":"AI Credits at $2.00 (global) or $2.20 (regional) per credit; features billed per million tokens, pages, or GB-month.","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Requires a Snowflake account; AI Credit spend is separate from warehouse spend and can be hard to budget."],"receipts":["https://www.snowflake.com/en/data-cloud/cortex","https://docs.snowflake.com/en/user-guide/snowflake-cortex/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"speechify","name":"Speechify","tagline":"Text-to-speech reader and AI voice generator for content consumption.","url":"https://speechify.com","category":"audio","pricing":"freemium","price_note":"free tier; Premium $29/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"The best-known reading assistant; natural voices make long articles and PDFs listenable. The AI voice-over studio is usable but not as expressive as ElevenLabs or Cartesia for character work.","limitations":["Best voices and features sit behind the yearly plan.","Voice-over output can sound like a single speaker regardless of content.","Privacy of uploaded documents needs review."],"strengths":["Natural voices make long articles and PDFs listenable.","Well-known reading assistant with broad platform availability.","AI voice-over studio handles basic content-conversion jobs."],"use_for":["Consuming long articles and PDFs as audio on the go.","Basic voice-over projects where expressiveness is not critical."],"skip_when":["Character work or expressive narration is required.","You need the best voices without paying for the yearly plan.","Privacy of uploaded documents is a core concern."],"receipts":["https://speechify.com","https://speechify.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"speechmatics","name":"Speechmatics","tagline":"Speech-to-text and text-to-speech APIs supporting 55+ languages with low latency and flexible deployment.","url":"https://www.speechmatics.com","category":"audio","pricing":"paid","price_note":"Speech-to-text starts at $0.24/hr for batch Standard and $0.43/hr for real-time Enhanced; text-to-speech is $0.011 per 1,000 characters.","free_tier":false,"open_source":false,"api":true,"self_host":true,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Real-time and batch pricing differ; private cloud, container, and on-device deployments require custom plans."],"receipts":["https://www.speechmatics.com","https://www.speechmatics.com/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"stable-diffusion","name":"Stable Diffusion (Stability AI)","tagline":"The open-weights classic — the largest community ecosystem of checkpoints and tools.","url":"https://stability.ai","category":"image","pricing":"freemium","price_note":"open weights free; Brand Studio Core $50/mo; API from ~$0.03/image","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"The value is the ecosystem — ControlNet, LoRAs, ComfyUI, thousands of community checkpoints. Base SDXL/SD3 quality lags behind FLUX now, but for custom pipelines and tutorials it still has the deepest bench.","limitations":["Base models aging against FLUX and newer rivals.","Company turbulence eroded trust in the roadmap.","Good results require picking the right community checkpoint, not the raw model."],"strengths":["The open-weights ecosystem gives you ControlNet, LoRAs, ComfyUI and thousands of checkpoints.","You can self-host pipelines and avoid per-image API metering.","Community tutorials and custom workflows are the deepest in the image category."],"use_for":["Building custom image pipelines where you need full control over the stack.","Self-hosted workflows that should not send assets to a third-party API.","Tutorials and experiments that rely on the largest community checkpoint library."],"skip_when":["You want the best base-model quality out of the box; FLUX currently leads.","You cannot spend time picking and tuning the right community checkpoint.","Vendor roadmap stability matters more than ecosystem breadth."],"receipts":["https://github.com/Stability-AI/generative-models"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"steel","name":"Steel","tagline":"Open-source cloud browsers for agents — sessions API, CAPTCHA, proxies, or Docker self-host.","url":"https://steel.dev","category":"data","pricing":"freemium","price_note":"Launch $0 + usage ($30 one-time credits, $0.10/browser-hour); Scale $250/mo + usage ($0.08/h, $100 credits/mo); proxy $6–$10/GB","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"Browserbase-shaped, with a public OSS docker story. Launch is $0 plus $0.10/hour after a $30 credit. Scale drops the hour rate and raises session length. CAPTCHA/proxies on Launch need a $10 paid deposit. Adjacent to Browser Use, not a duplicate.","limitations":["Launch sessions cap at 15 minutes; 24h sessions are Enterprise.","Proxies and CAPTCHA on Launch require a $10 paid balance, not free credits.","Stealth browser is Enterprise-only on the June 2026 limits table.","Browser hours round up to the next minute."],"strengths":["Open-source browser with a public Docker self-host path.","Sessions API with CAPTCHA solving and proxy support.","Launch tier starts at $0 plus usage-based browser hours."],"use_for":["Agent workflows that need real browser sessions instead of HTTP scraping.","Teams that want to self-host browser infra on their own hardware.","Prototypes that can fit within 15-minute Launch sessions."],"skip_when":["You need free CAPTCHA or proxy access without a $10 paid balance.","24-hour sessions or stealth browser are hard requirements.","You want fully predictable costs on heavy browser-hour volume."],"receipts":["https://docs.steel.dev/overview/pricinglimits","https://github.com/steel-dev/steel-browser"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"stepfun-api","name":"StepFun API","tagline":"API for Step-2/Step-3 family of frontier MoE models from Shanghai.","url":"https://platform.stepfun.com","category":"models-api","pricing":"paid","price_note":"Step 3.7 Flash ~$0.20/$1.15 per 1M; Step 3.5 Flash open-source $0.10/$0.30; Step-2 ~$0.83/$2.50; Step-2-mini ~$0.14/$0.42; Step-1.5V ~$0.42/$1.25; Step-R ~$1.11/$3.33","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","verdict":"situational","verdict_text":"Step 3.7 Flash is cheap, open weights, and scores well on coding and math benchmarks. Step-2 is a capable closed model at mid-tier prices. Brand recognition and docs outside China are weak, and access often runs through resellers.","limitations":["The hosted API is proprietary; Step 3.5 Flash and Step 3.7 Flash weights are separate open-weight releases.","StepFun runs separate China (platform.stepfun.com) and global (platform.stepfun.ai) platforms with non-interchangeable API keys.","Lower Western brand recognition; docs mostly Chinese.","Mixed access outside China; often reached via resellers.","Step-2 is closed, unlike Step 3.5 Flash and Step 3.7 Flash.","Smaller model family than Qwen or Tencent."],"strengths":["Step 3.7 Flash is cheap, open weights, and strong on coding and math.","Step-2 is a capable closed model at mid-tier prices.","The family covers text, vision, and reasoning variants."],"use_for":["Cost-sensitive coding and math workloads that fit the Flash SKUs.","Teams that want Step-2's closed-model quality without premium pricing."],"skip_when":["You need strong Western brand recognition and English-first docs.","You want one API key that works across China and global platforms.","You need a model family as broad as Qwen or Tencent."],"receipts":["https://platform.stepfun.ai","https://github.com/stepfun-ai/Step-3.7-Flash","https://apirank.vip/zh/tutorials/stepfun-api-review/","https://tokenmix.ai/blog/step-3-5-flash-review-stepfun-2026"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"stitch","name":"Stitch","tagline":"Google's experimental AI design tool for generating UI layouts and components.","url":"https://stitch.withgoogle.com","category":"design","pricing":"free","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Interesting if you want to see what a Google-backed AI design tool can generate, but it is still experimental. Output tends to need cleanup before it is production-ready.","limitations":["Experimental; availability and feature set change without warning.","Generated designs often need significant refinement.","Tied to Google's ecosystem and release schedule."],"strengths":["Google-backed experimental AI design tool for generating UI layouts and components.","Free to use while in Labs."],"use_for":["Quick explorations of AI-generated UI layouts and components.","Teams already inside Google's ecosystem and release schedule."],"skip_when":["You need production-ready designs without significant cleanup.","Availability and feature set must be stable.","You are not willing to wait on Google's release schedule."],"receipts":["https://stitch.withgoogle.com","https://blog.google/innovation-and-ai/models-and-research/google-labs/stitch-ai-ui-design/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"sudowrite","name":"Sudowrite","tagline":"Fiction-first AI writing — brainstorm, expand and revise novels.","url":"https://www.sudowrite.com","category":"writing","pricing":"paid","price_note":"from $10/mo (Hobby & Student)","free_tier":false,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"The only writing tool built by and for fiction writers. Story Engine, character bibles and scene expansion match how novels actually get written. Don't bother if you don't write fiction.","limitations":["Niche by design — useless for business writing.","Credit system is confusing.","Output still reads as AI fiction without heavy human revision."],"strengths":["Built by and for fiction writers, not a repurposed general writing tool.","Story Engine, character bibles, and scene expansion match how novels are written."],"use_for":["Fiction writers drafting novels, scenes, and character work.","Authors who want story-specific tools rather than generic prompts."],"skip_when":["You write business, marketing, or non-fiction content.","You cannot budget around an opaque credit system.","You are not prepared to heavily revise AI-sounding fiction output."],"receipts":["https://www.sudowrite.com","https://sudowrite.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"suno","name":"Suno","tagline":"Full-song generation — vocals, lyrics, arrangement — from a text prompt.","url":"https://suno.com","category":"audio","pricing":"freemium","price_note":"free tier (10 songs/day, non-commercial); Pro $10/mo, Premier $30/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"ship","verdict_text":"Type a prompt, get a full song with vocals in about a minute. Suno is the fastest way to mock up demos, background tracks or rough ideas. The legal situation around training data is still messy, commercial use needs a paid tier, and arrangement control is shallow. If you're just prototyping, it's the obvious choice.","limitations":["Training-data lawsuits keep a legal cloud over commercial use.","Commercial rights require a paid tier.","Fine control over arrangement and mix is limited.","Free tier capped at 7 lifetime downloads from 2026-09-03; Pro 20/mo, Premier 60/mo; ongoing label lawsuits/licensing deals."],"strengths":["Prompt to a full song with vocals in about a minute.","Fastest mock-up for demos, beds and rough ideas."],"use_for":["Prototype tracks and placeholder music in a product or film scratch.","Non-commercial sketching on the free tier while it still exists."],"skip_when":["Commercial use under the training-data lawsuits is a legal no.","You need mix-level control of arrangement.","You depend on free-tier downloads after 2026-09-03."],"receipts":["https://suno.com","https://suno.com/pricing","https://www.layer3labs.io/guides/suno-pricing","https://www.musicbusinessworldwide.com/suno-limits-subscribers-downloads-per-month/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"sunsama","name":"Sunsama","tagline":"Daily planner that unifies tasks and calendars so modern professionals can stay calm and focused.","url":"https://www.sunsama.com","category":"productivity","pricing":"paid","price_note":"No free forever plan; 14-day free trial, then paid subscription priced around $1/workday with a 20% annual discount.","free_tier":false,"open_source":false,"api":false,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","No free forever plan; after the 14-day trial the service requires a paid subscription."],"receipts":["https://www.sunsama.com","https://www.sunsama.com/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"supabase-mcp","name":"Supabase MCP","tagline":"Supabase's hosted MCP — SQL, migrations, functions. Vendor says keep it off production.","url":"https://supabase.com/docs/guides/ai-tools/mcp","category":"mcp","pricing":"free","price_note":"included with a Supabase account; hosted https://mcp.supabase.com/mcp; read_only=true and project_ref query params; branching tools need a paid plan","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"Official hosted MCP at mcp.supabase.com: list tables, execute SQL, apply migrations, deploy functions, even create projects. A read_only flag and project_ref exist. Supabase's own docs say never point this at production data — development and testing only.","limitations":["execute_sql and apply_migration are write tools unless read_only=true is set.","Unscoped URL can see every project in the org.","Storage group is off by default; branching is experimental and paid-plan."],"strengths":["Official Supabase server with OAuth (no PAT required for interactive clients).","Feature groups and read_only=true / project_ref on the URL, not a fork.","Local CLI MCP at http://localhost:54321/mcp for local development."],"use_for":["Agents scaffolding a new Supabase project or reading schema in development.","A read-only URL against a staging project, not the prod database."],"skip_when":["The project is production — vendor docs say this MCP is not for that.","You only needed docs search — that is a feature group, not a reason to enable SQL.","Branching tools on a Free plan — branching is a paid-plan group."],"receipts":["https://supabase.com/docs/guides/ai-tools/mcp","https://github.com/supabase/mcp"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":false},"last_verified":"2026-08-26","observed_by":"whysanesanders"},{"slug":"superhuman","name":"Superhuman","tagline":"Premium email client with AI triage, drafts and follow-ups.","url":"https://www.superhuman.com","category":"productivity","pricing":"freemium","price_note":"freemium suite; Free/Pro tiers cover Grammarly/Docs/Go but not Mail; Mail requires the Business suite: $40 per member/mo monthly or $33/mo billed annually ($396/yr)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Fast email if you are keyboard-driven and live in Gmail. The drafting is fine, but at $33–40/mo (Business tier, required for Mail) it is a luxury now that Gmail and others ship free AI. Try it for the speed, not because you need AI mail.","limitations":["Expensive while free alternatives catch up.","Calendar features thinner than dedicated tools.","Muscle-memory lock-in cuts both ways.","Mail is bundled only in Business/Enterprise suite tiers; Free/Pro tiers get Grammarly/Docs/Go, not Mail (the company formerly known as Grammarly rebranded as Superhuman in 2025)."],"strengths":["Keyboard-driven email speed for Gmail power users.","Fast drafting and follow-ups."],"use_for":["Gmail users who live on keyboard shortcuts.","Professionals prioritizing email speed over AI novelty."],"skip_when":["$33–40/mo Business tier for Mail is too expensive.","You rely on strong calendar features.","Free AI in Gmail meets your drafting needs."],"receipts":["https://www.superhuman.com","https://www.superhuman.com/plans","https://help.superhuman.com/hc/en-us/articles/46005733349517-Pricing-Plans"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"yes","local_processing":false},"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"synthesia","name":"Synthesia","tagline":"AI avatar video platform for corporate training and marketing.","url":"https://www.synthesia.io","category":"video","pricing":"paid","price_note":"from $29/mo (Starter)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"The default for templated talking-head video at scale. Avatars are presentation-ready, translations are decent, and the editor is genuinely usable by non-editors. Emotional range and gestural nuance still feel canned.","limitations":["Avatar expressiveness is limited compared to real presenters.","Pricing jumps quickly for multiple videos or custom avatars.","Output can feel samey if not carefully scripted."],"strengths":["Default for templated talking-head video at scale.","Avatars, translations and an editor non-editors can actually use."],"use_for":["Corporate training and localized talking-head updates.","Marketing explainers that should look presentation-ready, not cinematic."],"skip_when":["You need real presenters' emotional range.","Custom avatars and volume would jump the price.","Unscripted nuance is the point of the video."],"receipts":["https://www.synthesia.io","https://www.synthesia.io/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"tabby","name":"Tabby","tagline":"Self-hosted coding assistant: completions and an agent, runs on your GPU or a small box.","url":"https://tabby.tabbyml.com","category":"coding","pricing":"free","price_note":"Apache-2.0 self-host; you pay the hardware","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: we have not reproduced a Tabby deploy here.","Quality is the local model, not a Cursor-class bundled router."],"receipts":["https://tabby.tabbyml.com","https://github.com/TabbyML/tabby"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"tailark","name":"Tailark","tagline":"Marketing blocks and page kits built on shadcn/ui and Tailwind CSS.","url":"https://tailark.com","category":"design","pricing":"freemium","price_note":"free OSS kits; paid plans for Quartz blocks and pages","free_tier":true,"open_source":true,"api":false,"self_host":false,"verdict":"situational","verdict_text":"A strong choice for assembling marketing pages quickly. The free kits cover common sections, while the paid tier adds more polished page-level templates.","limitations":["The best blocks are paywalled.","Strongly tied to the shadcn/ui workflow.","Less useful outside React and Tailwind stacks."],"strengths":["Free open-source marketing blocks and page kits for quick assembly.","Built on shadcn/ui and Tailwind CSS, fitting that stack well.","Paid Quartz tier adds more polished page-level templates."],"use_for":["React/Tailwind marketing sites that need common sections fast.","Teams already using shadcn/ui who want prebuilt block libraries."],"skip_when":["You need premium blocks but cannot pay for the Quartz tier.","The project is outside React or Tailwind.","You want fully unique designs rather than shared block patterns."],"receipts":["https://tailark.com","https://github.com/tailark/blocks"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"tailwind-ui","name":"Tailwind Plus","tagline":"Official Tailwind CSS component library and templates.","url":"https://tailwindcss.com/plus","category":"design","pricing":"paid","price_note":"all-access from $299 one-time","free_tier":false,"open_source":false,"api":false,"self_host":false,"verdict":"ship","verdict_text":"The renamed Tailwind UI remains the gold standard for Tailwind marketing and application UI. The components are well-designed, accessible, and updated with each Tailwind release.","limitations":["One-time cost is significant for hobbyists.","No free tier to evaluate before buying.","Templates can look recognizable if used verbatim."],"strengths":["Gold standard for Tailwind marketing and application UI — well-designed and accessible.","Updated with Tailwind releases instead of a frozen 2022 kit."],"use_for":["A Tailwind product or marketing site that should not invent every block.","Teams that will pay once for official templates."],"skip_when":["A $299 one-time hit is too much for a hobby.","You need a free tier to evaluate before buying.","Verbatim templates would look like every other Tailwind Plus site."],"receipts":["https://tailwindcss.com/plus","https://tailwindcss.com/plus/changelog"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"talordata","name":"TalorData","tagline":"Real-time SERP API returning structured JSON from Google, Bing, Yandex, and DuckDuckGo.","url":"https://www.talordata.com","category":"data","pricing":"freemium","price_note":"500 free API responses on signup; then $0.25 per 1,000 responses pay-as-you-go","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Fast structured SERP data without proxy or CAPTCHA management — clean JSON from four major engines at a flat rate. Best fit when an agent or RAG pipeline needs raw search results; it returns ranked results, not synthesized research summaries like Tavily-style APIs.","limitations":["Free credits are one-time (500 on signup), not a recurring monthly tier.","API-only; no consumer-facing interface.","Fewer engines than SerpApi (no Baidu, Walmart, YouTube, etc.)."],"receipts":["https://www.talordata.com"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-26","observed_by":"talordata-serp-api"},{"slug":"tavily","name":"Tavily","tagline":"Search API optimized for LLM agents with structured, sourced answers.","url":"https://tavily.com","category":"data","pricing":"freemium","price_note":"free tier (1k calls/mo); paid from $30/mo","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"ship","verdict_text":"Purpose-built for retrieval-augmented generation: clean JSON, source citations, and fast response times. The default for agent search. Result quality drops on obscure technical queries.","limitations":["Free tier has tight rate limits.","Some obscure queries return thin results.","Advanced reranking requires paid plans."],"strengths":["Search API built for agents: clean JSON, citations, fast enough for RAG.","Default agent-search pick when you want structured, sourced hits."],"use_for":["Retrieval for LLM agents that must return sources.","Light production search on the paid plan after the free cap."],"skip_when":["Free-tier rate limits would be the production path.","The queries are obscure and come back thin.","You need paid reranking and will not pay from $30/mo."],"receipts":["https://tavily.com","https://tavily.com/pricing","https://docs.tavily.com/faq/faq"],"affiliate":"none","evidence_tier":"source-verified","guide":{"install":"Create a Tavily key and call the Search API. Most agent frameworks have a first-party Tavily tool.","requirements":["An API key","A job that needs sourced web snippets, not a full crawl"],"cost":"Free 1k calls/mo, then from $30/mo on the public list. Reranking sits on paid plans.","security":"Queries leave for Tavily. Do not put customer secrets in the search string.","breaks_when":["The query is obscure and the snippets come back thin.","You needed site-wide markdown — that is Firecrawl, not this."]},"momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"temporal","name":"Temporal","tagline":"Durable workflow orchestration engine, open-source and cloud-hosted, popular for reliable AI agent pipelines.","url":"https://temporal.io","category":"dev-infra","pricing":"freemium","price_note":"Open-source self-hosted under MIT license is free; Temporal Cloud starts at $100/mo Essentials and offers $1,000 in free credits.","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"situational","verdict_text":"Temporal turns fragile agent scripts and long-running AI pipelines into durable, observable workflows that survive crashes and retries. The open-source server is mature and MIT-licensed, while Temporal Cloud removes operational burden for a premium. It is not a low-code tool; teams must be willing to model workflows and activities in code.","limitations":["There is a learning curve: workers, workflows, activities, and event history are new concepts for many teams.","Self-hosted operation requires running the Temporal Server, persistence, and observability stack.","Cloud pricing scales with Actions and storage; high-throughput workloads need careful capacity planning."],"strengths":["Durable workflow orchestration that survives crashes and retries.","Mature MIT-licensed open-source server.","Observable long-running workflows out of the box."],"use_for":["Reliable AI agent pipelines that must survive failures.","Teams willing to model workflows in code for durability."],"skip_when":["You need a low-code automation tool.","The team is not willing to learn workers, workflows, and activities.","Cloud pricing for Actions and storage is unpredictable."],"receipts":["https://temporal.io","https://temporal.io/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"tencent-codebuddy","name":"Tencent CodeBuddy","tagline":"Tencent's AI coding suite with IDE, VS Code/JetBrains plugin, and CLI built on Hunyuan and DeepSeek.","url":"https://www.codebuddy.ai","category":"coding","pricing":"freemium","price_note":"Experience tier free (500 credits/mo); Standard ¥99/mo, Advanced ¥199/mo, Flagship ¥999/mo; Enterprise SaaS ¥198/seat/mo; credits shared with WorkBuddy","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["hunyuan-t1","deepseek-v4-pro"],"model_routing":"locked","verdict":"situational","verdict_text":"A strong regional option for Chinese developers who want plugin, IDE, and CLI forms under one subscription. The catch: credit-based pricing is hard to predict, the .ai marketing site is barely usable outside China, and non-Chinese users face language and billing barriers.","limitations":["Credit consumption depends on model and task complexity; monthly costs can surprise.","The international .ai site renders poorly; the Chinese .cn site is the practical source.","Enterprise SaaS requires seat minimums; private cloud starts at 100 seats.","Not open source; model routing limited to Tencent and partner models.","Documentation and support are Chinese-first."],"strengths":["IDE, VS Code/JetBrains plugin, and CLI forms under one subscription.","Built on Hunyuan and DeepSeek models for a regional Chinese audience.","Credit-based pricing can fit lighter usage if you watch consumption."],"use_for":["Chinese developers who want plugin, IDE, and CLI tools from one vendor.","Teams already in Tencent Cloud and comfortable with Chinese-first support."],"skip_when":["You need predictable credit pricing or tight cost controls.","You are outside China and rely on the .ai marketing site.","You need open-source tooling or broader model choice."],"receipts":["https://www.codebuddy.cn/pricing/","https://cloud.tencent.com/document/product/1831/134333","https://cloud.tencent.com/product/acc"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"tldv","name":"tl;dv","tagline":"AI-powered meeting notetaker that records, transcribes, and summarizes calls without a bot.","url":"https://tldv.io","category":"productivity","pricing":"freemium","price_note":"Free Forever plan includes unlimited recordings, transcriptions, and AI summaries in 30+ languages.","free_tier":true,"open_source":false,"api":false,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Free tier covers unlimited recordings and transcriptions; advanced team, CRM, and coaching features are gated behind paid plans."],"receipts":["https://tldv.io"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"no","local_processing":false},"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"todoist","name":"Todoist","tagline":"To-do list and task manager that helps professionals organize work and life with natural-language capture.","url":"https://todoist.com","category":"productivity","pricing":"freemium","price_note":"Beginner plan is free with 5 personal projects; Pro and Business plans add projects, filters, and team workspace features.","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Beginner free plan is limited to 5 personal projects, 3 filter views, and 1 week of activity history."],"receipts":["https://todoist.com","https://todoist.com/en/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"together-ai","name":"Together AI","tagline":"Cloud inference for open models — serverless endpoints and fine-tuning.","url":"https://www.together.ai","category":"dev-infra","pricing":"freemium","price_note":"starter credits; from $0.10/1M tokens (small open models)","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["llama-4-maverick","qwen3-8-max","deepseek-v4-flash"],"model_routing":"locked","verdict":"ship","verdict_text":"Solid uptime, broad catalog, serverless or dedicated endpoints. A safe place to move an open-weight prototype into production. Costs at scale push you toward dedicated GPUs, and the newest models sometimes land later than on specialist hosts.","limitations":["Costs at scale push you toward dedicated GPU contracts.","Newest models sometimes arrive later than at specialists.","Fine-tuning UX less polished than the inference side."],"strengths":["Solid uptime and a broad open-model catalog, serverless or dedicated.","Safe place to move an open-weight prototype into production."],"use_for":["Hosted open models without picking a GPU vendor yet.","Fine-tunes that can start on their inference side."],"skip_when":["Scale already points at dedicated GPU contracts.","You need the newest model the day a specialist hosts it.","Fine-tuning UX is the product, not an afterthought."],"receipts":["https://www.together.ai","https://www.together.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"tongyi-lingma","name":"Tongyi Lingma","tagline":"Free IDE plugin and standalone AI IDE from Alibaba for completion, chat and repo refactoring.","url":"https://lingma.aliyun.com","category":"coding","pricing":"free","price_note":"Completely free for individual developers; enterprise tier custom pricing via Alibaba Cloud","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"A genuinely free Copilot-style plugin with solid Chinese-language support. The catch is Alibaba ecosystem lock-in, limited English documentation, and weaker agentic depth than Cursor or Claude Code.","limitations":["Less powerful than Cursor or Claude Code for agentic tasks.","Heavy Alibaba ecosystem lock-in.","Limited English documentation and Western support.","Enterprise pricing is opaque and China-account oriented."],"strengths":["Genuinely free Copilot-style plugin for individual developers.","Solid Chinese-language support and a standalone AI IDE option.","Backed by Alibaba Cloud for China-account users."],"use_for":["Individual developers who want free completion and chat inside their IDE.","Chinese-speaking teams or projects already on Alibaba Cloud."],"skip_when":["You need agentic depth comparable to Cursor or Claude Code.","English documentation and Western support are required.","You cannot accept Alibaba ecosystem lock-in.","Transparent enterprise pricing outside China is essential."],"receipts":["https://www.ainspiro.com/en/news/china-ai-coding-tools-2026-lingma-comate-codegeex-trae/","https://gowukong.co/en/tools/tongyi-lingma/","https://recatools.com/ai-directory/tongyi-lingma/","https://www.stork.ai/ai-tools/tongyi-lingma"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"traceloop","name":"Traceloop","tagline":"LLM reliability platform that monitors output quality, traces requests, and runs evaluations in CI.","url":"https://www.traceloop.com","category":"dev-infra","pricing":"freemium","price_note":"Free forever for up to 50,000 spans per month and 5 seats; Enterprise is custom with on-premise and air-gapped deployment options.","free_tier":true,"open_source":false,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Free tier retains data for only 24 hours and is limited to 50,000 spans and 5 seats per month."],"receipts":["https://www.traceloop.com","https://www.traceloop.com/docs","https://www.traceloop.com/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"trae","name":"Trae","tagline":"ByteDance's AI-native IDE with inline completion, chat and autonomous SOLO agent mode.","url":"https://www.trae.ai","category":"coding","pricing":"freemium","price_note":"Free plan with monthly AI allowance; paid tiers Lite $3/mo, Pro $10/mo, Pro+ $30/mo, Ultra $100/mo (token-based); pricing changed 2026-02-18","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["claude-sonnet-5"],"model_routing":"bundled","verdict":"situational","verdict_text":"The cheapest way into a full AI IDE, and Lite at $3/mo undercuts everyone. The trade-off is ByteDance hosting your code with no BYOK option; fine for side projects, risky for proprietary work.","limitations":["ByteDance product — data and privacy concerns for proprietary code.","Token allowances can run out quickly on heavy agent sessions.","Closed source; no BYOK or standalone API.","Western documentation and support are thinner than Cursor/Windsurf."],"strengths":["Cheapest way into a full AI IDE: free plan plus Lite at $3/mo undercuts rivals.","Inline completion, chat, and autonomous SOLO agent mode in one editor.","Fast start for developers already used to VS Code-like layouts."],"use_for":["Side projects and budget-conscious developers who want an AI-native IDE.","Developers comfortable with ByteDance hosting for non-proprietary code."],"skip_when":["You are working on proprietary code and cannot accept ByteDance data handling.","You need BYOK, a standalone API, or audited closed-source tooling.","Western documentation and support quality are critical."],"receipts":["https://www.trae.ai","https://aipedia.wiki/tools/trae/","https://aiagentsquare.com/agents/trae","https://netizenschoice.com/trae-ai-review-2026"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"trigger-dev","name":"Trigger.dev","tagline":"Background jobs and workflows for TypeScript apps with AI-friendly primitives.","url":"https://trigger.dev","category":"dev-infra","pricing":"freemium","price_note":"self-host free; cloud from $10/mo (Hobby)","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"ship","verdict_text":"The best DX for long-running background jobs in TypeScript. Retries, scheduling, and observability are first-class. You need to be comfortable with TypeScript; other languages are not supported.","limitations":["TypeScript-only ecosystem.","Self-hosting needs infrastructure know-how.","Cloud pricing can surprise at high task volume."],"strengths":["Best DX for long-running background jobs in TypeScript: retries, schedules, observability.","Self-host is real; cloud is optional."],"use_for":["TypeScript apps that need durable AI and background work.","Teams that will self-host rather than surprise-bill at volume."],"skip_when":["The stack is not TypeScript.","Self-hosting ops is not in the team.","Cloud task volume would surprise the bill."],"receipts":["https://trigger.dev","https://github.com/triggerdotdev/trigger.dev"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"type","name":"Type","tagline":"AI-first collaborative writing app for notes, documents, and workflows.","url":"https://type.ai","category":"writing","pricing":"freemium","price_note":"free tier; Pro from $8/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"A clean, fast writing surface with inline AI that stays out of the way. Good for drafting and rewriting, but lacks the depth of Notion for databases or the publishing features of dedicated tools.","limitations":["Feature set is narrower than Notion/Obsidian.","Offline support is limited.","Collaboration features lag Google Docs."],"strengths":["Clean, fast writing surface with inline AI that stays out of the way.","Good for drafting and rewriting without tool overload."],"use_for":["Writers who want a minimal AI drafting surface, not a full workspace.","Quick rewriting and note-taking when database features don't matter."],"skip_when":["You need Notion-style databases or deep workspace features.","Offline availability or Google Docs-grade collaboration is required.","Publishing features are part of your workflow."],"receipts":["https://type.ai","https://type.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"typeface","name":"Typeface","tagline":"Enterprise brand-safe generative content platform (not a consumer writer).","url":"https://www.typeface.ai","category":"writing","pricing":"paid","free_tier":false,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: sales-led pricing, no public list we verified this week.","Enterprise brand vaults are the product; a solo founder does not need this."],"receipts":["https://www.typeface.ai"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"udio","name":"Udio","tagline":"High-fidelity music generation, strong on vocals and mix quality.","url":"https://www.udio.com","category":"audio","pricing":"freemium","price_note":"free tier; Standard $10/mo, Pro $30/mo (yearly)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Sounds a little better than Suno on some genres, and the extend/remix workflow is built for musicians. But downloads have been disabled since October 2025 because of label deals, the product direction is unclear, and full-song structure is less consistent. Try it, but don't buy a year up front.","limitations":["Downloads disabled since 2025-10 (UMG/Warner licensing deals) — you can't export your tracks.","Product direction uncertain after lawsuit settlements.","Less consistent than Suno on full-song structure.","Generation limits tight on lower tiers."],"strengths":["Strong vocal and mix quality on generated music in supported genres.","Extend/remix workflow fits musicians iterating on ideas, not just one-shot prompts."],"use_for":["Quick musical sketches and loops when you do not need a downloadable file.","Genres where Udio's vocal and mix quality sounds better than alternatives."],"skip_when":["You need downloadable tracks for release, distribution, or offline use.","Full-song structure and consistency are required.","You want a stable, predictable product direction after the label deals."],"receipts":["https://www.udio.com","https://www.udio.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"redshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"uizard","name":"Uizard","tagline":"AI-powered UI design tool for wireframes and mockups.","url":"https://uizard.io","category":"design","pricing":"freemium","price_note":"free tier; paid plans from $12/creator/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Handy for non-designers who need quick wireframes from sketches or prompts. The output is good enough for early ideation, not final design handoff.","limitations":["Generated designs lack polish for production use.","Free tier is limited in projects and exports.","Collaboration features are basic compared to Figma."],"strengths":["Turns sketches or prompts into wireframes without design expertise.","Fast early ideation for non-designers and product generalists."],"use_for":["Rapid wireframing in discovery before a designer gets involved.","Non-designers who need a visual artifact to communicate an idea."],"skip_when":["You need production-polish fit for final design handoff.","Free-tier project and export limits would block the workflow.","Team collaboration needs exceed basic Figma-level workflows."],"receipts":["https://uizard.io","https://uizard.io/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"unstructured","name":"Unstructured","tagline":"Enterprise platform that turns complex unstructured files into clean, AI-ready structured data.","url":"https://unstructured.io","category":"data","pricing":"freemium","free_tier":true,"open_source":true,"api":true,"self_host":true,"limitations":["Radar: not field-run in this catalog wave.","Managed platform pricing is not shown on the homepage; open-source library requires self-hosted infrastructure."],"receipts":["https://unstructured.io"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"upstage","name":"Upstage","tagline":"Enterprise AI platform with Solar LLM, document parsing, and information extraction APIs.","url":"https://upstage.ai","category":"models-api","pricing":"freemium","price_note":"Every agent includes 10 free runs; prepaid commitment tiers start at $100/month; per-page and per-step billing.","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"locked","limitations":["Radar: not field-run in this catalog wave.","Pricing is credit-based with prepaid commitments, and purchased credits are generally non-refundable."],"receipts":["https://upstage.ai","https://www.upstage.ai/pricing"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"upstash-vector","name":"Upstash Vector","tagline":"Serverless vector database with dense, sparse, and hybrid indexes via REST API or SDKs.","url":"https://upstash.com/vector","category":"data","pricing":"freemium","price_note":"Free plan includes 10,000 queries and 10,000 updates daily; paid tiers include Pay-as-you-go, Fixed, Pro, and Enterprise.","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: not field-run in this catalog wave.","Eventually consistent; newly upserted vectors may have a short delay before querying."],"receipts":["https://upstash.com/vector"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"v0","name":"v0","tagline":"Vercel's AI design-to-code tool for React and Tailwind components.","url":"https://v0.app","category":"design","pricing":"freemium","price_note":"Free $0/mo with $5 monthly credits; Plus $30/user/mo; Business $100/user/mo; usage consumes model-priced credits","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Turns a prompt into a React/Tailwind component you can copy and tweak. Handy for a quick landing section or dashboard shell, but generated code usually needs cleanup for accessibility, responsive edge cases, and production conventions.","limitations":["Generated code often ignores accessibility and edge cases.","Credit limits can interrupt flow on the free tier.","Heavily favors Vercel and Next.js conventions."],"strengths":["Generates React/Tailwind components from a text prompt.","Faster than starting from scratch for simple UI blocks."],"use_for":["Scaffolding a landing section or dashboard shell to iterate on.","Teams already using Next.js and Vercel conventions."],"skip_when":["You need accessible, production-ready code out of the box.","Free-tier credit limits would interrupt your flow.","Your stack is not Next.js/Vercel-friendly."],"receipts":["https://v0.app","https://v0.app/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"v0-dev","name":"v0","tagline":"AI generator of full Next.js/React apps from a prompt, screenshot, or Figma frame, with one-click Vercel deploy.","url":"https://v0.app","category":"coding","pricing":"freemium","price_note":"Free $0/mo ($5 credits/mo, 7 messages/day); Plus $30/user/mo; Business $100/user/mo; Premium $20 retired for new users. Generations billed by tokens (v0 Pro ~$2/1M input, $10/1M output).","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Turns a prompt, screenshot or Figma frame into a working Next.js/React/Tailwind/shadcn app — from landing pages to full-stack with API routes, DB and auth. Preview in the Vercel Sandbox and one-click deploy. Best for a Vercel-hosted MVP or internal tool inside the React stack; outside that stack you are paying to rewrite everything, and agent iterations can burn credits on failed attempts before the code is production-ready.","limitations":["Free tier is demo-grade (7 messages/day, $5 credits).","Output is Next.js/React/Tailwind/shadcn-first — other stacks need rework.","Agent iterations can burn credits on failed attempts; production code still needs review."],"strengths":["Generates full Next.js/React/Tailwind/shadcn apps from prompts, screenshots, or Figma frames.","One-click deploy to Vercel and preview in the Vercel Sandbox.","Covers landing pages through full-stack with API routes, DB, and auth."],"use_for":["Vercel-hosted MVPs and internal tools inside the React stack.","Rapid prototypes from a Figma frame or screenshot."],"skip_when":["Your stack is not Next.js/React/Tailwind/shadcn-first.","You cannot tolerate unpredictable token-credit burn on failed agent iterations.","Free tier limits (7 messages/day, $5 credits) are too demo-grade for the task."],"receipts":["https://v0.app/pricing","https://v0.app/docs/faqs"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"veo-3","name":"Veo 3.1","tagline":"Google's video model with native audio generation — the current quality leader.","url":"https://deepmind.google/models/veo/","category":"video","pricing":"freemium","price_note":"free daily credits in Flow; Google AI Pro $19.99/mo (1000 Flow credits/mo); Ultra 5x $99.99/mo (10k Flow credits/mo) / 20x $199.99/mo (25k Flow credits/mo)","free_tier":true,"open_source":false,"api":true,"self_host":false,"models_used":["veo-3-1-generate"],"verdict":"ship","verdict_text":"The current quality leader. Physics, lip-sync and native audio are ahead of the pack right now. Per-second pricing is steep, so use it for final shots and short deliverables, not for rapid iteration.","limitations":["Per-second API pricing is steep for experimentation.","Access tiers and quotas across Gemini/Flow/Vertex are confusing.","Roughly 8-second clip limit per generation.","Veo 3.0 shutdown is marked \"coming soon\" on Google's deprecations page (no date set); no standalone Veo subscription (via Google AI plans/Flow)."],"strengths":["Current quality leader: physics, lip-sync and native audio are ahead of the pack.","Right tool for final short shots, not for burning credits on sketches."],"use_for":["Hero shots and short deliverables where native audio matters.","Gemini/Flow users who already have the credits."],"skip_when":["Per-second API pricing would make experimentation the expensive part.","You need clips longer than ~8 seconds in one generation.","Access across Gemini, Flow and Vertex is too confusing for the deadline."],"receipts":["https://deepmind.google/models/veo/","https://ai.google.dev/pricing","https://ai.google.dev/gemini-api/docs/deprecations?hl=en","https://one.google.com/about/ai-premium"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":true,"last_verified":"2026-09-01","observed_by":"whysanesanders"},{"slug":"vercel-ai-sdk","name":"Vercel AI SDK","tagline":"TypeScript toolkit for AI apps — unified providers, streaming UI primitives.","url":"https://ai-sdk.dev","category":"dev-infra","pricing":"free","price_note":"free, open source","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"ship","verdict_text":"The standard glue for AI features in TypeScript apps. One interface across OpenAI, Anthropic, Google and many others, plus streaming primitives and React hooks that behave sensibly. The abstraction leaks when you need provider-specific features, and the docs churn with each release.","limitations":["The abstraction leaks on advanced provider-specific features.","Docs and examples churn with fast releases.","TypeScript/JavaScript only."],"strengths":["Standard TypeScript glue across OpenAI, Anthropic, Google and others.","Streaming primitives and React hooks that behave like product UI, not a demo."],"use_for":["Next.js or React apps that should stream tokens from more than one lab.","A unified provider layer before you invent your own."],"skip_when":["You need advanced provider-specific features the abstraction leaks.","Docs churn with each release would stall the team.","The app is not TypeScript or JavaScript."],"receipts":["https://github.com/vercel/ai"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"vidu","name":"Vidu","tagline":"Text/image/reference-to-video generator with native audio and strong character consistency.","url":"https://www.vidu.com","category":"video","pricing":"freemium","price_note":"free plan includes 80 credits/mo plus unlimited off-peak generation; paid plans credit-based; API credits $0.005 each but generations priced per operation by model/resolution/duration (Q2-turbo 540P from $0.03, Q2-Pro 1080P from $0.425)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Strong reference-to-video consistency and a cheap entry point for creators. The aesthetic leans cinematic/anime, the API pricing matrix is complex, and the vendor is smaller and China-based.","limitations":["Aesthetic leans cinematic/anime/narrative, less suited for UGC or e-commerce ads.","API pricing matrix is complex.","Smaller vendor with China-based operations.","Western access and support are thinner than Runway or Kling."],"strengths":["Strong reference-to-video consistency for characters.","Cheap entry point with free credits plus unlimited off-peak generation.","Native audio generation baked into the video output."],"use_for":["Cinematic or anime-style narrative videos.","Projects that lean on reference images for consistent characters."],"skip_when":["You need UGC or e-commerce ad aesthetics.","The complex API pricing matrix is a procurement barrier.","You need strong Western vendor support and access."],"receipts":["https://www.vidu.com/pricing","https://www.vidu.com/blog/ai-image-animator-free","https://platform.vidu.com/docs/pricing","https://aisotools.com/pricing/vidu-ai"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"visily","name":"Visily","tagline":"AI wireframing and prototyping tool for non-designers — prompt, screenshot, or sketch to editable mockups.","url":"https://www.visily.ai","category":"design","pricing":"freemium","price_note":"Starter $0 (300 AI credits/mo, limited boards/elements, watermarked exports); Pro $11/editor/mo annually ($14 monthly); Business $29/editor/mo annually; Enterprise custom","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"Fastest way for product managers, founders, and non-designers to turn a prompt or screenshot into an editable wireframe or prototype. The catch: the free tier is capped and watermarked, paid seats are per-editor, and high-fidelity polish still needs a designer or Figma handoff.","limitations":["Free Starter has limited boards/elements and watermarked 1x exports.","AI credits are consumed per action; heavy AI use burns through monthly allowances quickly.","Per-editor pricing can outpace small-team budgets.","Prototype fidelity is wireframe-to-mid-fi; production-ready visual design usually requires Figma."],"strengths":["AI wireframing and prototyping for non-designers from prompt or screenshot.","Fast path for product managers and founders to editable mockups.","Sketch-to-editable-wireframe conversion lowers the design starting barrier."],"use_for":["Product managers and founders turning ideas into shareable wireframes.","Teams that need mid-fi prototypes without hiring a designer first."],"skip_when":["You need production-ready, high-fidelity visual design in the same tool.","Free watermarked exports and credit caps break your workflow.","Per-editor pricing would outpace a small team's budget."],"receipts":["https://www.visily.ai","https://www.visily.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"vllm","name":"vLLM","tagline":"High-throughput OSS inference engine — PagedAttention, OpenAI-compatible server, you bring GPUs.","url":"https://github.com/vllm-project/vllm","category":"dev-infra","pricing":"free","price_note":"Apache-licensed engine free; you pay GPUs/cloud; no vLLM seat fee","free_tier":true,"open_source":true,"api":true,"self_host":true,"model_routing":"byok","verdict":"situational","verdict_text":"The default way to serve open weights if you already have NVIDIA (or a growing list of other accelerators). OpenAI-compatible HTTP, continuous batching, quantization zoo. The bill is hardware and the on-call, not a SaaS SKU. Not a hosted model API. Not field-run in this pass.","limitations":["You operate CUDA/ROCm, drivers, and capacity; this is not serverless inference.","Day-0 support for a brand-new architecture can lag the lab's own stack.","Multi-node parallelism is real ops, not a checkbox.","No public vLLM Cloud price — third-party hosts are other cards."],"strengths":["Default open-source engine for serving open weights at high throughput.","OpenAI-compatible HTTP server with continuous batching and quantization.","No SaaS seat fee; the bill is your hardware and on-call time."],"use_for":["Self-hosted open-weight inference on NVIDIA or supported accelerators.","Teams that need quantization and batching control under their own stack."],"skip_when":["You want serverless inference without operating GPUs and drivers.","Day-0 model support or multi-node parallelism would block you.","You cannot staff the on-call needed for a self-hosted inference layer."],"receipts":["https://github.com/vllm-project/vllm","https://docs.vllm.ai/en/latest/"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"voiceflow","name":"Voiceflow","tagline":"Collaborative builder for support/voice/chat agents with a design-tool UX.","url":"https://www.voiceflow.com","category":"automation","pricing":"freemium","free_tier":true,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: no field run of current pricing or runtime.","Prototype-to-prod gap is the usual failure mode of canvas builders."],"receipts":["https://www.voiceflow.com"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"voidleap-code","name":"Voidleap Code","tagline":"Local-first agentic IDE with your own keys, per-turn model switching, and full observability.","url":"https://voidleap.com/","category":"coding","pricing":"freemium","price_note":"Free: 3 workspaces, 3 projects each, 1 key profile, 7-day history; Pro €20/seat/mo; Team €50/seat/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"byok","verdict":"situational","verdict_text":"Positions itself as a harness rather than a wrapper: you see system prompts, context entries, and cost/latency charts for every thread. Bring-your-own-key model support is broad, including Anthropic, OpenAI, Gemini, Ollama, OpenRouter, and Bedrock. The free plan is real but small; most solo developers will outgrow the project limits quickly.","limitations":["Pro and Team tiers are still in development; only the Free plan ships today.","macOS and Windows desktop only.","Requires periodic online license validation even though the runtime is local."],"strengths":["Full observability with system prompts, context entries, and cost/latency charts.","Broad bring-your-own-key support across Anthropic, OpenAI, Gemini, Ollama, OpenRouter and Bedrock.","Local-first IDE that keeps the runtime on your machine."],"use_for":["Developers who want per-turn model switching and cost visibility.","Teams that need to use their own API keys instead of a bundled vendor pool."],"skip_when":["You need shipped Pro or Team tiers today.","Linux or non-desktop workflows are required.","Periodic online license validation conflicts with fully offline use."],"receipts":["https://voidleap.com/","https://voidleap.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"voila","name":"Voila","tagline":"Cross-platform AI assistant — chat, writing and automation from any tab or app.","url":"https://www.getvoila.ai","category":"productivity","pricing":"freemium","price_note":"free tier; Premium $10/mo ($8 yearly, 3k requests/mo); Ultimate $20/mo ($16 yearly, unlimited)","free_tier":true,"open_source":false,"api":false,"self_host":false,"verdict":"situational","verdict_text":"A friendly-priced assistant that genuinely runs everywhere — browser, desktop and mobile. Premium at $8/mo billed yearly is one of the cheaper honest entry points in this category. The catch is tier gating: Premium locks you to the Auto model with a request cap, and model choice only arrives with Ultimate. Solid as one convenient assistant; power users will hit the ceiling.","limitations":["Premium is limited to the Auto model — no model choice on that tier.","Premium caps at 3,000 requests per month."],"strengths":["Cross-platform assistant running in browser, desktop, and mobile.","Premium at $8/mo yearly is a cheap honest entry point.","Ultimate tier unlocks model choice."],"use_for":["Users who want one convenient assistant across every device.","Budget buyers who can live with the Auto model and 3k request cap."],"skip_when":["You need model choice below the Ultimate tier.","3,000 requests per month would cap a Premium workflow.","You are a power user who wants full control over the underlying model."],"receipts":["https://www.getvoila.ai/pricing"],"affiliate":"declared","affiliate_url":"https://www.getvoila.ai/?via=noemium","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"voyage-ai","name":"Voyage AI","tagline":"Embedding and rerank APIs — large free-token grants, then cheap per-million rates.","url":"https://www.voyageai.com","category":"data","pricing":"freemium","price_note":"voyage-4 $0.06/1M after 200M free tokens; voyage-4-lite $0.02/1M; rerank-2.5 $0.05/1M after 200M free; Batch API 33% off, no free-token apply","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Retrieval people already know the name. The public docs (updated 2026-08-13) still give 200M free tokens on voyage-4 family, then $0.02–$0.12/1M. Rerank is a separate meter. Batch is cheaper and burns no free grant. Not a vector database — pair with Chroma/Qdrant/Pinecone.","limitations":["Free tokens do not apply to Batch API.","Domain models (finance/law/code-2) only include 50M free tokens.","Multimodal is token + pixel billed; tiny images still hit a 50k-pixel floor.","Fine-tuned dedicated instances are sales-only."],"strengths":["Generous 200M free-token grant on the voyage-4 family.","Cheap per-million rates after the free grant.","Strong name recognition in retrieval and rerank stacks."],"use_for":["RAG pipelines that need embeddings and reranking from one vendor.","Teams building retrieval stacks with Chroma, Qdrant, or Pinecone.","Workloads that can use batch mode to burn no free grant and save cost."],"skip_when":["You planned to use the free-token grant for Batch API calls.","You need domain-specific models beyond the 50M free allocation.","You want self-serve fine-tuned dedicated instances rather than sales-only."],"receipts":["https://docs.voyageai.com/docs/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"wan","name":"Wan","tagline":"Alibaba Tongyi Wanxiang video models — hosted on tongyi.aliyun.com/wan; older weights on GitHub.","url":"https://tongyi.aliyun.com/wan","category":"video","pricing":"freemium","price_note":"Tongyi web app + Alibaba Cloud DashScope; hosted video is regional pay-as-you-go (confirm in console — no stable public USD card fetched); Wan2.1 Apache weights are self-hostable","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"This is Wan the video stack, not Qwen-Image. The consumer surface is Tongyi Wanxiang; API is DashScope. GitHub Wan2.1 is the open-weights generation, not a promise that the latest hosted Wan is OSI. China-region console and moderation apply. No USD list price we could pin without the billing console.","limitations":["No durable public USD price table fetched; treat Alibaba console as source of truth.","Latest hosted Wan is not the same artifact as Apache Wan2.1 weights.","China-region account, payment, and content-moderation constraints.","Not Qwen-Image (that's the still-image card)."],"strengths":["Alibaba-hosted video generation through Tongyi Wanxiang.","Wan2.1 Apache weights are available for self-hosting.","API access via DashScope for programmatic use."],"use_for":["Video generation projects that can work within Alibaba Cloud's region.","Teams that want to self-host Wan2.1 open weights.","Users already inside the Tongyi ecosystem."],"skip_when":["You need a stable public USD price card without console access.","China-region account, payment or moderation constraints are blockers.","You expected the latest hosted Wan to match the open Wan2.1 weights."],"receipts":["https://github.com/Wan-Video/Wan2.1","https://tongyi.aliyun.com/wan"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"warp","name":"Warp","tagline":"Modern terminal with a first-party coding agent — credits for Warp Agent, BYOK on Free.","url":"https://www.warp.dev","category":"coding","pricing":"freemium","price_note":"Free $0 (BYOK + limited agent); Build from $20/mo (1,500 credits); Max from $200/mo (18,000 credits); Business from $50/user/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"model_routing":"byok","verdict":"situational","verdict_text":"The terminal people already wanted, now with an agent meter. Free stays a real terminal plus BYOK; Build is $20 for 1,500 credits at API-shaped rates. Factories (cloud agent fleets) are a second SKU with a 20% PAYG markup. Closed client. No hands-on run in this catalog pass.","limitations":["Agent credits and terminal UX are sold together; heavy agent use is not the Free plan.","Cloud agents and Factories are extra surface area with their own markup.","Closed source; you cannot audit the client or self-host the agent service.","Team SSO and BYOLLM sit on Business/Enterprise, not Build."],"strengths":["Modern terminal UX plus AI agent in one tool.","Free tier keeps a real terminal and BYOK.","Public credit math: $20 for 1,500 credits."],"use_for":["Developers who want terminal modernization plus AI in one app.","Free-tier users happy with BYOK for agent tasks.","Teams needing cloud agent fleets via Factories."],"skip_when":["Heavy agent use would force expensive Build/Max tiers.","You need an auditable open-source terminal.","Team SSO/BYOLLM requirements rule out Build."],"receipts":["https://www.warp.dev/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"weaviate","name":"Weaviate","tagline":"Open-source vector database with hybrid search and modular ML integrations.","url":"https://weaviate.io","category":"data","pricing":"freemium","price_note":"self-host free; cloud free tier, Flex from $45/mo","free_tier":true,"open_source":true,"api":true,"self_host":true,"verdict":"ship","verdict_text":"The most feature-complete open vector DB: hybrid BM25+vector search, multi-tenancy, and solid clients. Self-hosts cleanly, and the managed cloud is optional rather than a trap.","limitations":["Memory-hungry at scale.","Operational complexity grows with sharding.","Docs span too many product versions."],"strengths":["Most feature-complete open vector DB: hybrid BM25+vector, multi-tenancy, solid clients.","Self-hosts cleanly; managed cloud is optional rather than a trap."],"use_for":["RAG that needs hybrid search, not vectors alone.","Multi-tenant apps that should stay on an open engine."],"skip_when":["Memory at scale would dominate the box.","Sharding ops would outgrow the team.","Docs spanning too many versions would slow you down."],"receipts":["https://github.com/weaviate/weaviate"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"wellsaid","name":"WellSaid","tagline":"Studio TTS for enterprise voiceovers, older than the ElevenLabs wave, still sold into learning/ops.","url":"https://wellsaidlabs.com","category":"audio","pricing":"paid","free_tier":false,"open_source":false,"api":true,"self_host":false,"limitations":["Radar: no current price page reproduced this week.","If you wanted character/emotion, Hume/ElevenLabs are the comparison set, not this card as a verdict."],"receipts":["https://wellsaidlabs.com"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"whisper","name":"Whisper","tagline":"OpenAI's open-weights speech recognition — the default transcription model.","url":"https://github.com/openai/whisper","category":"audio","pricing":"free","price_note":"open source (MIT), self-host free; hosted: gpt-4o-transcribe $0.006/min","free_tier":true,"open_source":true,"api":true,"self_host":true,"models_used":["whisper-v3"],"verdict":"ship","verdict_text":"The main reason paid transcription is optional now. large-v3 is accurate across dozens of languages and runs on a laptop. It still hallucinates on silence and noisy segments, doesn't do speaker diarization out of the box, and needs a decent GPU for real-time. Use it as the baseline for any transcription pipeline — hosted or self-hosted.","limitations":["Hallucinates text on silence and noisy segments.","No built-in speaker diarization (pair with pyannote).","large-v3 needs a decent GPU for real-time work.","whisper-1 API is legacy; the current hosted line is the gpt-transcribe family."],"strengths":["The reason paid transcription is optional: large-v3 is accurate across dozens of languages.","Open weights that run on a laptop; hosted gpt-transcribe exists when you do not want ops."],"use_for":["A baseline STT pipeline, hosted or self-hosted.","Offline or private transcription where a vendor STT is the wrong boundary."],"skip_when":["Silence and noise would hallucinate text you cannot filter.","You need speaker diarization without pairing pyannote.","Real-time large-v3 would run without a decent GPU."],"receipts":["https://github.com/openai/whisper","https://developers.openai.com/api/docs/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"whisperx","name":"WhisperX","tagline":"Open-source speech recognition toolkit with batched Whisper, word-level timestamps, and speaker diarization.","url":"https://github.com/m-bain/whisperX","category":"audio","pricing":"free","free_tier":true,"open_source":true,"api":false,"self_host":true,"model_routing":"bundled","limitations":["Radar: not field-run in this catalog wave.","Overlapping speech is not handled well and speaker diarization is far from perfect, per the project's own limitations list."],"receipts":["https://github.com/m-bain/whisperX"],"affiliate":"none","evidence_tier":"radar","momentum":"blueshift","featured":false,"last_verified":"2026-09-01","observed_by":"catalog-radar"},{"slug":"windsurf","name":"Devin Desktop (ex Windsurf)","tagline":"Agentic IDE (VS Code fork) with Devin agents and cloud coding.","url":"https://devin.ai","category":"coding","pricing":"freemium","price_note":"free tier; Pro $20/mo, Max $200/mo","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["claude-sonnet-5","gpt-5-6-terra"],"model_routing":"bundled","verdict":"situational","verdict_text":"A credible Cursor alternative with a cleaner agent UX. But it's on its second owner in a year — the OpenAI deal collapsed, Cognition bought it, Cascade got retired. Worth a look when Cursor's limits bite, not before.","limitations":["Rebranded to Devin Desktop after the Cognition acquisition; Cascade retired 2026-07-01.","Smaller extension and community ecosystem than Cursor.","Ownership turbulence (failed OpenAI deal, Cognition acquisition) makes the roadmap uncertain.","Agent occasionally loops on large, unfamiliar repos."],"strengths":["Cleaner agent UX than Cursor in some workflows.","Credible fallback when Cursor's token limits or pricing bite."],"use_for":["Teams hitting Cursor limits who want an alternative agent IDE.","Users curious about a different agent experience in a VS Code fork."],"skip_when":["You need a stable long-term roadmap from a settled owner.","Large, unfamiliar repos are common and agent looping would hurt.","A smaller extension ecosystem would stall your workflow."],"receipts":["https://devin.ai","https://devin.ai/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"wordtune","name":"Wordtune","tagline":"AI rewrite/compose assistant from AI21, aimed at emails and docs rather than long-form novels.","url":"https://www.wordtune.com","category":"writing","pricing":"freemium","free_tier":true,"open_source":false,"api":false,"self_host":false,"limitations":["Radar: no field run. Browser-extension privacy not verified here.","Not a replacement for a model API you already pay for."],"receipts":["https://www.wordtune.com"],"affiliate":"none","evidence_tier":"radar","momentum":"steady","featured":false,"last_verified":"2026-08-26","observed_by":"catalog-radar"},{"slug":"workato","name":"Workato","tagline":"Enterprise integration and automation platform with AI-assisted workflow building.","url":"https://www.workato.com","category":"automation","pricing":"paid","price_note":"enterprise pricing","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"The enterprise-safe choice for connecting SaaS and on-prem systems. Security and governance are strong, but pricing is opaque and the product is overkill for small teams.","limitations":["No transparent self-serve pricing.","Heavy UI for simple automations.","Implementation timelines can stretch for complex integrations."],"strengths":["Enterprise-grade security and governance for SaaS and on-prem integrations.","Strong connector coverage for large, regulated environments."],"use_for":["Enterprises that need governed integrations across SaaS and on-prem systems.","Teams with implementation budget for complex, multi-system workflows."],"skip_when":["You need transparent self-serve pricing before evaluating.","The automation is simple and would be buried by a heavy UI.","Implementation timelines must be short and predictable."],"receipts":["https://www.workato.com","https://docs.workato.com"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"writer","name":"Writer","tagline":"Enterprise AI platform for on-brand marketing, writing, and agentic workflows.","url":"https://writer.com","category":"writing","pricing":"paid","price_note":"Starter and Enterprise per-seat plans; no self-serve free tier (Enterprise includes limited Lite seats)","free_tier":false,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Writer targets large marketing and content teams that need brand governance, not casual writers. Playbooks and style guides keep outputs consistent, but the platform is expensive and overkill for individuals. Enterprise seat model with no real free tier.","limitations":["No self-serve free plan; paid per-seat entry with enterprise sales motion.","Best fit is regulated/enterprise marketing teams, not general writing.","Integrations and setup require org-level configuration before it pays off."],"strengths":["Strong brand governance through playbooks and style guides.","Built for large marketing and content teams, not casual writers.","Keeps outputs on-brand across enterprise workflows."],"use_for":["Regulated/enterprise marketing teams that need consistent brand voice.","Organizations with the ops capacity to configure integrations and setup."],"skip_when":["You are an individual or small team without enterprise budget.","You need a self-serve free tier to evaluate first.","You lack the org-level configuration bandwidth to make it pay off."],"receipts":["https://writer.com","https://writer.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"writesonic","name":"Writesonic","tagline":"AI search visibility and SEO agentic platform — track brand mentions, fix citations, prove ROI.","url":"https://writesonic.com","category":"writing","pricing":"freemium","price_note":"7-day free trial; paid from $79/mo billed annually (Starter)","free_tier":true,"open_source":false,"api":true,"self_host":false,"verdict":"situational","verdict_text":"Combines article writing, SEO scoring, and chatbot creation in one dashboard, which is convenient if you want a single visibility tool. The $79/mo Starter plan is not cheap for generic SEO scores, and every generated article needs a fact-check before it ships.","limitations":["SEO suggestions are generic; human expertise still wins.","Generated articles need fact-checking and rewriting.","Interface can feel cluttered with upsells."],"strengths":["All-in-one dashboard for AI articles, SEO scoring, and chatbot creation.","7-day free trial before committing to the $79/mo Starter annual plan.","Tracks brand mentions and citations for AI search visibility."],"use_for":["Content marketers who need draft production plus basic SEO scoring.","Teams that want a single visibility dashboard instead of point tools.","Trying AI search visibility with a no-card free trial."],"skip_when":["You need expert-level SEO recommendations rather than generic scores.","You cannot dedicate time to fact-check every generated article.","A clean UI without constant upsells is a hard requirement."],"receipts":["https://writesonic.com","https://writesonic.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"xai-api","name":"xAI API","tagline":"Grok models via API — strong reasoning, X integration, big context.","url":"https://platform.x.ai","category":"models-api","pricing":"paid","price_note":"pay-as-you-go; Grok 4.6 $2/$6 per 1M","free_tier":false,"open_source":false,"api":true,"self_host":false,"models_used":["grok-4-6"],"model_routing":"locked","verdict":"situational","verdict_text":"Grok 4 keeps up on benchmarks and the API is straightforward. The ecosystem is young, though, and the X association is a real procurement risk for some teams. Pricing isn't disruptive compared with Claude or GPT.","limitations":["Young platform: fewer SDKs, tools and battle-testing.","Pricing is not disruptive versus Claude or GPT.","Reputation risk some orgs will not accept."],"strengths":["Grok 4 keeps up on benchmarks with a straightforward API.","Big context window and strong reasoning on paper.","Tight X integration for teams already on that platform."],"use_for":["Projects that want Grok models through a simple API.","Teams comfortable with the X ecosystem tie-in.","Use cases where benchmark-reported reasoning matters."],"skip_when":["The X association is a procurement risk your org will not accept.","You need a mature ecosystem with many SDKs and integrations.","Pricing must undercut Claude or GPT to justify the switch."],"receipts":["https://platform.x.ai","https://docs.x.ai/docs"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"yi-api","name":"01.AI Yi API","tagline":"Self-serve Yi API platform is shutting down; open-weight Yi models remain on HuggingFace.","url":"https://platform.lingyiwanwu.com","category":"models-api","pricing":"paid","price_note":"yi-lightning ¥0.99/1M (~$0.14) and yi-vision-v2 ¥6/1M (~$0.85), 16K context; new signups/top-ups closed 2026-08-03, API calls end 2026-09-03","free_tier":false,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"skip","verdict_text":"01.AI is winding down the self-serve Yi API platform. A 2026-08-03 notice stopped new signups and top-ups; API calls end 2026-09-03 and balance refunds run through 2026-12-03. The open-weight Yi models remain available, but the hosted API is no longer a viable choice for new projects.","limitations":["The self-serve API platform is shutting down: new signups/top-ups closed 2026-08-03; API calls stop 2026-09-03.","Existing balances can be refunded through 2026-12-03; contact yi@01.ai.","The current card is RMB-only and requires Chinese-platform identity verification; the international surface closed in 2024.","yi-lightning and yi-vision-v2 are intelligent-routing SKUs; the actual model served may be DeepSeek or Qwen, not Yi.","Open-weight Yi models are separate from the hosted API and still free on HuggingFace.","The company is pivoting to sales-led enterprise platforms (TrueNorth/Wanzhi); no public pricing for those.","Self-serve API closes 2026-09-03 (announced 2026-08-03): registrations and top-ups stopped, existing calls work until then — enterprise sales only after."],"strengths":["yi-lightning is cheap on a per-token basis.","Open-weight Yi models remain free on HuggingFace.","Shutdown timeline is clearly published."],"use_for":["Existing users winding down before 2026-09-03.","Experimenting with open-weight Yi models on HuggingFace."],"skip_when":["You need a hosted API for a new project.","Chinese-platform identity verification is a blocker.","You want stable SKU identity rather than intelligent routing."],"receipts":["https://platform.lingyiwanwu.com","https://www.usagepricing.com/blueprint/01-ai","https://nolist.ai/item/yi-lightning"],"affiliate":"none","evidence_tier":"source-verified","momentum":"redshift","featured":false,"last_verified":"2026-08-24","observed_by":"whysanesanders"},{"slug":"zapier","name":"Zapier","tagline":"The automation incumbent — 7000+ app integrations, now with AI steps.","url":"https://zapier.com","category":"automation","pricing":"freemium","price_note":"free 100 tasks/mo; Professional from $19.99/mo billed yearly (750 tasks) or $29.99/mo monthly","free_tier":true,"open_source":false,"api":true,"self_host":false,"model_routing":"bundled","verdict":"situational","verdict_text":"Still unbeatable for connecting standard SaaS apps reliably. But AI workflows eat tasks fast — every LLM call costs you — and n8n or Make handle AI-native automation for less. Keep it for plain plumbing, not agent loops.","limitations":["Per-task pricing explodes once AI loops enter workflows.","AI features feel bolted on vs AI-native rivals.","Complex branching becomes unmaintainable in the UI."],"strengths":["7000+ app integrations and reliable SaaS plumbing.","AI steps extend classic automations without leaving the platform.","Still the default for connecting standard SaaS apps reliably."],"use_for":["Plain SaaS-to-SaaS automations where reliability beats AI-native cost.","Teams already invested in Zapier's integration catalog."],"skip_when":["Your workflow is an AI agent loop that would burn through task pricing.","You need AI-native automation at lower cost than per-task LLM calls.","Complex branching would become unmaintainable in the visual UI."],"receipts":["https://zapier.com/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"steady","featured":false,"last_verified":"2026-08-19","observed_by":"whysanesanders"},{"slug":"zcode","name":"ZCode","tagline":"Agentic IDE from Z.ai/Zhipu built around GLM-5.3 for goal-driven coding with multi-agent workflows.","url":"https://zcode.z.ai","category":"coding","pricing":"freemium","price_note":"app download free; GLM Coding plan Lite ¥94.4/mo, Pro ¥430.4/mo, Max ¥862.4/mo (discounted from ¥118/¥538/¥1078); pay-as-you-go model API also available on BigModel.cn","free_tier":true,"open_source":false,"api":false,"self_host":false,"models_used":["glm-5-3","glm-5-2"],"model_routing":"locked","verdict":"situational","verdict_text":"A practical pick for developers already inside the Z.ai/GLM ecosystem who want a dedicated agentic IDE with predictable monthly credits. The catch: model choice is effectively GLM-only, the Linux build is still in beta, and the IDE itself is not open source despite some third-party claims.","limitations":["Not open source; no public source repository could be verified.","Model selection is tightly coupled to the GLM family and partner models.","Pricing and billing are presented in Chinese yuan with unclear international payment support.","Linux support is marked Beta on the downloads page.","Requires a Z.ai/BigModel.cn account and credits; BYOK is not prominently documented."],"strengths":["Dedicated agentic IDE built around GLM-5.3 for goal-driven coding.","Multi-agent workflows inside a single editor.","Free app download with predictable monthly credit plans.","Pay-as-you-go model API available via BigModel.cn."],"use_for":["Developers already inside the Z.ai/GLM ecosystem.","Teams that want predictable monthly credits for agentic IDE use."],"skip_when":["You require open-source or auditable tooling.","You need model choice beyond the GLM family and partner models.","Linux is your primary OS; the build is still marked Beta.","You need clear international payment support or documented BYOK."],"receipts":["https://zcode.z.ai","https://bigmodel.cn/pricing"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"last_verified":"2026-08-20","observed_by":"whysanesanders"},{"slug":"zed","name":"Zed","tagline":"Fast open-source editor (Rust) with a built-in AI panel and agent mode.","url":"https://zed.dev","category":"coding","pricing":"freemium","price_note":"editor free; Pro $10/mo includes $5 token credits, overage at API list price +10%; Business $30/seat/mo; BYOK free","free_tier":true,"open_source":true,"api":false,"self_host":true,"models_used":["claude-sonnet-5"],"model_routing":"byok","verdict":"situational","verdict_text":"Probably the fastest editor around, and the agent panel is honest engineering, not a skin. Bring your own key, point it at any provider — including the cheap Chinese ones — and it gets interesting. Still rougher than Cursor for heavy agent sessions.","limitations":["Plugin ecosystem far smaller than any VS Code fork.","Hosted AI features moved behind accounts and paid tiers over time.","Windows support historically lags macOS/Linux."],"strengths":["Probably the fastest editor around, built in Rust.","Honest engineering in the AI panel, not a thin skin.","Bring your own key to any provider, including cheaper Chinese APIs."],"use_for":["Developers who want speed and a native-feeling editor.","Users who prefer BYOK and transparent model choice.","Teams comfortable with a smaller but growing plugin ecosystem."],"skip_when":["You depend on the full VS Code extension marketplace.","You need the most polished heavy-agent experience today.","Windows parity is a hard requirement right now."],"receipts":["https://zed.dev/pricing","https://github.com/zed-industries/zed"],"affiliate":"none","evidence_tier":"source-verified","momentum":"blueshift","featured":false,"data_sensitivity":{"trains_on_inputs":"unknown","local_processing":true},"last_verified":"2026-08-24","observed_by":"whysanesanders"}]}