Up to four models
o1 vs DeepSeek-R1
On the index (GPT-4 = 100), o1 scores 136 and DeepSeek-R1 130. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
o1 leads DeepSeek-R1 by 6 points, ahead on 1 of 2 shared core tests.
Test by testthird-party results
o1
DeepSeek-R1
Every result as a table
| Test | o1 | DeepSeek-R1 |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 76.8% | 71.7% |
| Fiction.LiveBench | 83.3% | 69.4% |
| SimpleBench | 40.1% | 30.9% |
| Maths | ||
| AIME-style maths (OTIS mock) | 73.3% | 53.3% |
| MATH (level 5) | 94.7% | 93.0% |
| Coding | ||
| Aider Polyglot | 61.7% | 56.9% |
| WeirdML | 46.1% | 36.5% |
Task length · METR31 sec
31 sec
Tests at least two of these models have taken. Bold rows are the core tests.