GPT OSS 20B: API pricing, providers and specs
GPT OSS 20B is a text model by OpenAI. 26 providers serve it, from $0.09 to $0.55 /1M; the lowest price is at OpenRouter.
A newer version is out: GPT OSS 120B · released Aug 4, 2025
How GPT OSS 20B compares
Against the 737 text models on the site priced the same way.
- Cheap for its class: at $0.09 /1M, its lowest price undercuts 94% of comparable models (median $1.10).
- Its 131K-token context window is larger than 21% of them (median 262K), about 200 pages of text.
- On LMArena's Text board it ranks #86 of the 120 comparable models scored there.
- 26 providers serve it; 93% of comparable models have fewer.
- Its weights are open under the Apache 2.0 license: commercial use allowed.
GPT OSS 20B prices by provider
Every provider's list price, read from its own price page or API. A price marked “from” changes with context length, resolution or mode.
| Provider | Input | Output | Prompt caching | Context |
|---|---|---|---|---|
 Nvidia | Free/1M | Free/1M | — | — |
| OpenRouter | $0.018/1M | From$0.09/1M | $0.009/1M | 131K |
 Hugging Face | $0.03/1M | From$0.14/1M | — | — |
| Vercel AI Gateway | $0.03/1M | $0.14/1M | — | 131K |
 Deep Infra | $0.03/1M | From$0.14/1M | — | — |
 Poe | $0.03/1M | $0.141/1M | — | 131K |
 Kenari | $0.035/1M | $0.149/1M | $0.004/1M | — |
 Cortecs | $0.045/1M | $0.168/1M | — | — |
 SiliconFlow | $0.04/1M | $0.18/1M | — | — |
 Helicone | $0.05/1M | From$0.20/1M | — | — |
| AI/ML API | $0.05/1M | $0.20/1M | $0.04/1M | — |
 Jiekou.AI | $0.05/1M | $0.20/1M | — | — |
 Vertex | $0.07/1M | From$0.25/1M | $0.007/1M | — |
 FastRouter | $0.07/1M | From$0.25/1M | $0.007/1M | 131K |
| Groq | $0.075/1M | $0.30/1M | — | — |
 Cloudflare Workers AI | $0.20/1M | $0.30/1M | — | — |
 Prime Inference | $0.07/1M | $0.30/1M | — | — |
 Requesty | $0.07/1M | From$0.30/1M | $0.035/1M | 131K |
 Amazon Bedrock | $0.07/1M | From$0.30/1M | — | — |
 Impossibl | $0.075/1M | $0.30/1M | $0.037/1M | — |
 CometAPI | $0.08/1M | From$0.32/1M | — | — |
| Replicate | $0.09/1M | $0.36/1M | — | — |
 LaoZhang API | $0.10/1M | $0.40/1M | — | — |
 DigitalOcean | $0.05/1M | $0.45/1M | — | — |
| DevPass (LLM Gateway) | $0.10/1M | From$0.50/1M | — | — |
 AIHubMix | $0.11/1M | $0.55/1M | — | 128K |
- OpenRouterFrom$0.09
- Vercel AI Gateway$0.14
- Hugging FaceFrom$0.14
- Deep InfraFrom$0.14
- Poe$0.141
- Kenari$0.149
- Cortecs$0.168
- SiliconFlow$0.18
- AI/ML API$0.20
- Jiekou.AI$0.20
- HeliconeFrom$0.20
- VertexFrom$0.25
Specs
- Type
- Text
- Maker
- OpenAI
- Inputs
- Text
- Outputs
- Text
- Context
- 131K≈ 200 pages of text
- Max output
- 33K
- Capabilities
- JSON, Reasoning, Tool calling
- Released
- Aug 4, 2025
- License
- Apache 2.0commercial use allowed
- Open weights
- Yes, downloadable
- Parameters
- 21B
- Arena score
- 1287Text
- Monthly searches
- 12K
3 uncensored versions · See them
Alternatives to GPT OSS 20B
Same kind of model from other makers, at a similar price.
| Model | Lowest price | Context | Arena | Providers |
|---|---|---|---|---|
| GPT OSS 20Bthis model | $0.09/1M | 131K | 1287 | 26 |
 DeepSeek V4.1 Flash | $0.03/1M | 1.05M | 1664 | 39 |
 GLM-5.3-Flash | $0.20/1M | 1.05M | 1608 | 45 |
| DeepSeek V4 Flash | $0.17/1M | 1.05M | — | 26 |
 Ling 3.1 Flash | $0.225/1M | 262K | — | 5 |
 Qwen2-1.5B-Instruct | $0.20/1M | — | — | 2 |
 MiMo-V2.6-Flash | $0.10/1M | 1.05M | 1634 | 22 |
| Qwen3.5 9B | $0.13/1M | 262K | — | 12 |
| Mercury 2.5 | $0.15/1M | 260K | — | 5 |
More from OpenAI
GPT OSS 20B FAQ
The lowest price for GPT OSS 20B is $0.09 /1M, at OpenRouter.
OpenRouter, at $0.09 /1M. 26 providers serve GPT OSS 20B; the table above lists every one with its price.
GPT OSS 20B reads up to 131K tokens per request and writes up to 33K.
Yes. GPT OSS 20B's weights can be downloaded from its maker's Hugging Face page, under the Apache 2.0 license.
About 21B, as the safetensors metadata of its official Hugging Face repo states.
26 API providers serve GPT OSS 20B. Compare their prices above, or open a provider to see everything it offers.
Can't find what you're looking for? Contact us