gpt-oss-120bOpen Weight
Released
gpt-oss-120b is an AI model released by OpenAI on Tuesday, Aug 5 2025, 56 days after o3-pro. It is an open-weight model — the trained weights are available to download and run. It is a 117B parameter model with a 128k token context window. Benchmark results (shown below) cover MMLU, BullshitBench v2, SWE-Bench Verified, Humanity's Last Exam, and GPQA Diamond.
Available from
| AkashML | — | — | $0.03 | $0.17 | 128K | bf16 |
|---|---|---|---|---|---|---|
| CoreWeave | — | — | $0.03 | $0.17 | 128K | fp4 |
| DeepInfra | — | — | $0.037 | $0.17 | 128K | bf16 |
| Novita | — | — | $0.05 | $0.25 | 128K | fp4 |
| DigitalOcean | — | — | $0.055 | $0.385 | 128K | — |
| Mancer 2 | — | — | $0.06 | $0.50 | 128K | fp8 |
| BaseTen | — | — | $0.10 | $0.50 | 128K | fp4 |
| DeepInfra | — | turbo | $0.15 | $0.60 | 128K | bf16 |
| Amazon Bedrock | — | — | $0.15 | $0.60 | 128K | — |
| Amazon Bedrock | eu-west-1 | — | $0.15 | $0.60 | 128K | — |
| Groq | — | — | $0.15 | $0.60 | 128K | — |
| Nebius | — | — | $0.15 | $0.60 | 128K | fp4 |
| Phala | — | — | $0.15 | $0.60 | 128K | — |
| Together | — | — | $0.15 | $0.60 | 128K | — |
| Parasail | — | — | $0.10 | $0.75 | 128K | fp4 |
| Mara | — | — | $0.15 | $0.75 | 128K | — |
| Cerebras | — | — | $0.35 | $0.75 | 128K | fp16 |
| SambaNova | — | — | $0.14 | $0.95 | 128K | — |
Benchmarks
Coding
Reasoning & science
Robustness
About
gpt-oss-120b, released August 5, 2025 alongside the smaller gpt-oss-20b, was OpenAI's first open-weight language model since GPT-2 in 2019 — six years in which the company that popularised the term "open" shipped nothing downloadable. Both models came under Apache 2.0 with no usage restrictions and no commercial gate. The 120B is a mixture-of-experts design with about 5.1B parameters active per token, sized to fit on a single 80GB accelerator, while the 21B sibling was built to run on a high-end laptop.
The launch figures put it close to OpenAI's own mid-tier closed models: 80.1% on GPQA Diamond and 62.4% on SWE-Bench Verified, 90.0% on MMLU, and 14.9% on Humanity's Last Exam without tools, rising to 19.0% with them. OpenAI framed the release as a response to the Chinese open-weight wave — DeepSeek R1 and Kimi K2 had spent the preceding months setting the pace for freely downloadable models — and shipped it two days before GPT-5, which took the attention.
Compare gpt-oss-120b with
Suggested comparisons
Frequently asked questions
gpt-oss-120b was released by OpenAI on Tuesday, Aug 5 2025.
gpt-oss-120b was built by OpenAI. Creators of ChatGPT and the GPT series of models. Pioneered large-scale language model research.
gpt-oss-120b reports 6 tracked benchmark scores — BullshitBench v2: 11%; SWE-Bench Verified: 62.4%; Humanity's Last Exam (no tools): 14.9%; Humanity's Last Exam (with tools): 19%; GPQA Diamond: 80.1%; MMLU: 90%. Scores are the figures published at release by OpenAI.
gpt-oss-120b has a context window of 128k. That is the maximum amount of input plus output the model can hold in a single request.
gpt-oss-120b is reported at 117B parameters.
Partly. gpt-oss-120b is an open-weight model: the trained weights are free to download and run locally or on your own infrastructure, but the training data and code are not fully released and the license may restrict some commercial uses. It is not open source in the strict sense.
OpenAI's previous tracked release was o3-pro on Jun 10 2025, 56 days earlier. It was followed by gpt-oss-20b on Aug 5 2025.