{"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":"voyage-ai","canonical_url":"https://noemium.com/tools/voyage-ai/","attribution":"Noemium catalog data is licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). See https://noemium.com/method/ for how entries are verified."}