OpenAI
o1-preview benchmark results
5 results on 5 benchmarks from 3 sources.
Closed Released: Sep 12, 2024 Updated Oct 22, 2026
Also listed as: o1-preview-2024-09-12 · W&B Programmer O1 crosscheck5
All benchmark results
| Benchmarks | Score | Rank | Setting | Source type | Date |
|---|---|---|---|---|---|
| LMArena Text | 1389 | 176 | Overall, style control on; 31122 votes; 95% CI 1384–1394 | Official leaderboard | Oct 8, 2026 |
| GPQA Diamond | 50.3% | 204 | Epoch AI run, accuracy | Independent run | Jan 27, 2025 |
| SWE-bench Verified | 64.6% | 40 | agent: W&B Programmer O1 crosscheck5 | Official leaderboard | Jan 17, 2025 |
| AIME (OTIS Mock AIME 2024–2025) | 31.1% | 190 | Epoch AI run, OTIS Mock AIME 2024–2025 | Independent run | Mar 7, 2025 |
| MATH Level 5 | 81.7% | 35 | Epoch AI run, accuracy | Independent run | Jan 27, 2025 |
FAQ
Where do these o1-preview scores come from?
From Epoch AI Benchmarking Hub, LMArena leaderboard dataset, and SWE-bench leaderboards (swebench.com). Each row links to the original leaderboard, dataset or report, and vendor-reported numbers are labelled.