Up to four models
o3 vs DeepSeek-R1-0528
On the index (GPT-4 = 100), o3 scores 147 and DeepSeek-R1-0528 135. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
o3 leads DeepSeek-R1-0528 by 12 points, ahead on 2 of 2 shared core tests.
Test by testthird-party results
o3
DeepSeek-R1-0528
Every result as a table
| Test | o3 | DeepSeek-R1-0528 |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 81.8% | 76.3% |
| Fiction.LiveBench | 88.9% | 75.0% |
| SimpleBench | 53.1% | 40.8% |
| Maths | ||
| AIME-style maths (OTIS mock) | 84.4% | 66.4% |
| MATH (level 5) | 97.8% | 96.6% |
| Coding | ||
| Aider Polyglot | 81.3% | 71.4% |
| WeirdML | 52.4% | 41.6% |
| Agents | ||
| DeepResearch Bench | 45.2% | 35.1% |
Task length · METR39 sec
32 sec
Tests at least two of these models have taken. Bold rows are the core tests.