GPT-OSS
OpenAI·
LLMsopen-weightlocal
Overview
Open source release of a highly capable model.
Capabilities and innovations
Open weightsCommunity fine-tuningTransparency
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 76.2% | unsourced |
| Humanity’s Last Exam (no tools) | 18.9% | unsourced |
| SWE-bench Verified | 45.1% | unsourced |
| MMLU | 87.5% | unsourced |
- SWE-bench Pro: not reported by the vendor
- MMLU-Pro: not reported by the vendor
- Terminal-Bench 2.x: benchmark did not exist at release
- MMMU-Pro: benchmark did not exist at release
Links
More from OpenAI
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.