Up to four models
o4-mini vs Gemini 2.5 Pro
On the index (GPT-4 = 100), o4-mini scores 141 and Gemini 2.5 Pro 141. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
o4-mini and Gemini 2.5 Pro are too close to call (0 points apart, within the uncertainty), ahead on 0 of 2 shared core tests.
Test by testthird-party results
o4-mini
Gemini 2.5 Pro
Every result as a table
| Test | o4-mini | Gemini 2.5 Pro |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 79.6% | 83.8% |
| Humanity's Last Exam | 18.1% | 18.2% |
| Fiction.LiveBench | 77.8% | 66.7% |
| GeoBench | 64.0% | 78.0% |
| SimpleBench | 38.7% | 51.6% |
| VPCT | 57.5% | 48.0% |
| Maths | ||
| MATH (level 5) | 97.8% | 95.6% |
| Coding | ||
| Aider Polyglot | 72.0% | 72.9% |
| CadEval | 62.0% | 64.0% |
Task length · METR38 sec
Tests at least two of these models have taken. Bold rows are the core tests.