Hunyuan API
situationalTencent Cloud TokenHub API for Hunyuan Hy3, image, video and 3D models (now Tencent Hy).
Visit cloud.tencent.com →> **Hunyuan API** — situational > Verified 2026-08-24 > Limitation: The hosted API is proprietary; Hy3 weights are released under Apache 2.0 and can be run separately. > > https://noemium.com/tools/hunyuan-api/Add to kitOpen in NoemiumFact-check this catalog entry for Hunyuan API (models-api). 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. 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https://www.atlascloud.ai/providers/tencent https://www.apiuspro.cn/tutorial/tencent Last verified: 2026-08-24 Find outdated prices, wrong claims, or missing limitations. Cite primary sources (docs, pricing pages, benchmarks, model cards). Do not use marketing pages.[](https://noemium.com/tools/hunyuan-api/)
ledger entry · observed by @whysanesanders
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.
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.
Known 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.
Use it when
- Cost-sensitive Chinese API access for text and multimodal generation.
- Bulk asset generation when you accept ByteDance's trust boundary on BytePlus.
Skip it 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.
Buy-check
- ·Price — paid, 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 as of 2026-08-24. Open the pricing page, not a screenshot.
- ·Free tier — Free tier exists — burn it first.
- ·Cancel fallout — What happens to your data/projects on cancel — check the ToS, not the FAQ.
- ·Seat math — Per-seat pricing multiplies quietly. Count seats before the annual plan.
- ·Who skips it — Also skipped when: You need clean USD rate cards or straightforward Western signup.; EU/UK/South Korea geo-blocking is a blocker for your users..
Price trail — 2 revisionschangelog →
- 2026-W35Hy3 ~$0.14/$0.56 per 1M tokens (256K context) on Toke…Hy3 ~$0.14/$0.56 per 1M tokens (256K context) on Toke…
- 2026-W34Hunyuan TurboS ~$0.11/$0.28 per 1M (256K context); T1…Hy3 ~$0.14/$0.56 per 1M tokens (256K context) on Toke…
Facts
- pricing
- paid1
- 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
- yes — paid to keep using
- open source
- no
- api
- yes
- self-host
- no
- category
- models-api
- evidence
- source-verified
- Models
- locked
Sources
- 1cloud.tencent.com/document/product/1729/97731[snapshot]
- 2cloud.tencent.com/developer/article/2674909[snapshot]
- 3atlascloud.ai/providers/tencent[snapshot]
- 4apiuspro.cn/tutorial/tencent[snapshot]
No source, no number.
Similar joball alternatives →
| tool | verdict | price | oss | self-host |
|---|---|---|---|---|
| Hunyuan API | situational | paid | no | no |
| Google AI Studio / Gemini API | ship | freemium | no | no |
| Anthropic API | ship | paid | no | no |
| Cerebras | ship | freemium | no | no |
Facts from the catalog files — not scores you can buy. Quality and speed stay in the briefing, not in a fake 1–5 grid.
Google AI Studio / Gemini API
Gemini API with a real free tier — huge context, multimodal, generous limits.
APIFREE TIER
freemiumfree tier (text models); paid from $0.75/1M input (Gemini 3.7 Flash promo till 2027-01)
Anthropic API
Claude models direct — the practitioner's pick for coding and long-form text.
API
paidpay-as-you-go, from $1/1M input (Haiku 4.5); small free credits for new users
Cerebras
Ultra-fast inference on wafer-scale AI accelerators with competitive API pricing.
APIFREE TIER
freemiumfree trial ($5 credits); self-serve Developer from $10 deposit; Enterprise custom