Model tracker / OpenAI
gpt-oss-20b
gpt-oss-20b is a fast open-weight model from OpenAI, released 5 August 2025. At Google it costs $0.07 per million input tokens and $0.25 per million output tokens, with a 131K-token context window. 5 providers serve it, and Google is the cheapest.
- Made by
- OpenAI
- Released
- 2025-08-05
- Status
- Generally available
- Tier
- Fast
- Weights
- Open
- Context window
- 131,072 tokens
- Max output
- 32,768 tokens
- Input
- text
- API id
gpt-oss-20b-maas
gpt-oss-20b price at 5 providers
US dollars per million tokens. Blended is 3 input tokens for every output token, one number to compare on. The cheapest blended price is highlighted.
| Provider | Input | Output | Cached input | Blended | Batch | Checked |
|---|---|---|---|---|---|---|
| $0.07 | $0.25 | $0.007 | $0.115 | 50% off | 2026-10-14 | |
| AWS · openai.gpt-oss-20b-1:0 | $0.07 | $0.30 | – | $0.128 | 50% off | 2026-10-14 |
| Groq · openai/gpt-oss-20b | $0.075 | $0.30 | $0.037 | $0.131 | 50% off | 2026-10-14 |
| Replicate · openai/gpt-oss-20b | $0.09 | $0.36 | – | $0.158 | – | 2026-10-14 |
| Cloudflare · @cf/openai/gpt-oss-20b | $0.20 | $0.30 | – | $0.225 | – | 2026-10-14 |
- Google: Vertex AI.
- AWS: Amazon Bedrock, US East (N. Virginia), Standard tier. Flex and Batch 50% off; Priority $0.1225 / $0.525.
- Cloudflare: Billed in Neurons at $0.011 per 1,000; 10,000 Neurons a day free.
Models like gpt-oss-20b
Other fast models closest in price, at each maker's own price or the cheapest provider.
| Model | Input | Output | Context | Weights |
|---|---|---|---|---|
| gpt-oss-20b | $0.07 | $0.25 | 131K | Open |
| Qwen3.5 9B Alibaba | $0.10 | $0.15 | 262K | Open |
| Ministral 3 3B Mistral AI | $0.10 | $0.10 | 256K | Open |
| gpt-oss-120b OpenAI | $0.05 | $0.25 | 131K | Open |
| Ministral 3 8B Mistral AI | $0.15 | $0.15 | 256K | Open |
| Llama 3.3 70B Meta | $0.10 | $0.32 | 128K | Open |
Compare them all in the model tracker →
gpt-oss-20b FAQ
How much does gpt-oss-20b cost?
gpt-oss-20b costs $0.07 per million input tokens and $0.25 per million output tokens at Google, and $0.007 per million cached input tokens, with 50% off for batch requests. Prices are list prices in US dollars, last checked 14 October 2026.
What is gpt-oss-20b's context window?
gpt-oss-20b takes up to 131,072 tokens in one request and writes up to 32,768 tokens in one response.
Is gpt-oss-20b open weight?
Yes. OpenAI publishes the weights, so you can run gpt-oss-20b yourself or at any host that serves it.
Where can I use gpt-oss-20b?
gpt-oss-20b is available from Google, AWS, Groq, Replicate and Cloudflare. At Google, call it as gpt-oss-20b-maas.