AI Rate Cardgpt-oss API pricing: gpt-oss-120b and gpt-oss-20b at every provider
Updated Nov 1, 2026
gpt-oss is OpenAI's family of open-weight models under the Apache 2.0 license. OpenAI publishes the weights rather than selling them through its own API, so gpt-oss is sold by inference clouds and API resellers - over forty providers sell gpt-oss-120b, which makes it one of the most widely sold models in the catalog. Both sizes are mixture-of-experts models that reason before answering. Below: the typical price (the median across providers) and live tables of every seller. The tables are rebuilt from today's prices on every page view.
The models in one list
Price per 1M tokens, input / output, in the 1 November 2026 collection; "typical" is the median across providers:
- gpt-oss-120b - typically $0.15 / $0.60, from $0.03 / $0.10; 131K context, 41 providers. The larger model, about 5B parameters active per token.
- gpt-oss-20b - typically $0.05 / $0.20, from $0.01 / $0.05; 131K context, 25 providers. Small enough for cheap high-volume jobs.
- gpt-oss-safeguard-120b and gpt-oss-safeguard-20b - typically $0.15 / $0.60 and $0.07 / $0.21, 6 providers each. Versions tuned to classify content against a policy you write.
- A few providers also give gpt-oss away for free, usually with rate limits - see free AI APIs.
gpt-oss-120b: every provider
| Provider | Model | Input, $ | Output, $ |
|---|---|---|---|
| fastrouter | OpenAI: gpt-oss-120b (Free) | Free | Free |
| kilo | OpenAI: gpt-oss-120b | $0.027 | $0.098 |
| openrouter | OpenAI: gpt-oss-120b (batch) | $0.0296 | $0.14 |
| orcarouter | GPT OSS 120B | $0.03 | $0.17 |
| trustedrouter | OpenAI: gpt-oss-120b | $0.03165 | $0.18 |
| requesty | GPT OSS 120B | $0.032 | $0.14 |
| kenari | GPT OSS 120B | $0.03474 | $0.19 |
| abacus | GPT-OSS 120B | $0.037 | $0.17 |
| deepinfra | GPT OSS 120B | $0.037 | $0.17 |
| edenai | GPT OSS 120B | $0.037 | $0.17 |
| fastrouter | OpenAI: GPT OSS 120B | $0.037 | $0.17 |
| huggingface | GPT OSS 120B | $0.037 | $0.17 |
| helicone | OpenAI GPT-OSS 120b | $0.04 | $0.16 |
| edenai | GPT OSS 120B | $0.05 | $0.25 |
| novita | GPT OSS 120B | $0.05 | $0.25 |
Live data: this table is built from today's prices.
gpt-oss-20b: every provider
| Provider | Model | Input, $ | Output, $ |
|---|---|---|---|
| fastrouter | OpenAI: gpt-oss-20b (Free) | Free | Free |
| nvidia | GPT OSS 20B | Free | Free |
| edenai | GPT OSS 20B | $0.009 | $0.045 |
| kilo | OpenAI: gpt-oss-20b | $0.01 | $0.04 |
| trustedrouter | OpenAI: gpt-oss-20b | $0.01 | $0.04748 |
| edenai | GPT OSS 20B | $0.02 | $0.10 |
| openrouter | OpenAI: gpt-oss-20b (batch) | $0.024 | $0.11 |
| deepinfra | GPT OSS 20B | $0.03 | $0.14 |
| edenai | GPT OSS 20B | $0.03 | $0.14 |
| fastrouter | OpenAI: GPT OSS 20B | $0.03 | $0.14 |
| huggingface | GPT OSS 20B | $0.03 | $0.14 |
| openrouter | OpenAI: gpt-oss-20b | $0.03 | $0.13 |
| vercel | GPT OSS 20B | $0.03 | $0.14 |
| poe | GPT OSS 20B | $0.0303 | $0.14 |
| kenari | GPT OSS 20B | $0.03474 | $0.15 |
Live data: this table is built from today's prices.
How to choose
- The stronger model: gpt-oss-120b - with forty-plus sellers the spread is wide, from $0.03 to well above the median for input, so compare before you pick.
- The lowest price: gpt-oss-20b, about a third of the 120b price at the median.
- Moderation: gpt-oss-safeguard - it reads your policy and labels the content.
- Other cheap open models: the cheapest LLM APIs, Gemma, Qwen.
- Every gpt-oss offer in one list: gpt-oss in the catalog.
All prices are in USD per 1 million tokens, without taxes, as the providers publish them; each row links to its card with the source and the date it was checked.