OpenAI o1
OpenAI·
LLMsreasoningcloud
Overview
Reasoning model designed to spend more time thinking before responding.
Capabilities and innovations
Chain of thoughtComplex math/codingSelf-correctionChain of Thought ReasoningTest-time Compute ScalingSelf-correction via RL
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 78% | unsourced |
| Humanity’s Last Exam (no tools) | 15.2% | unsourced |
| SWE-bench Verified | 48.9% | unsourced |
| MMLU | 91.8% | unsourced |
| MMLU-Pro | 83.49% | unsourced |
- SWE-bench Pro: benchmark did not exist at release
- 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.