All Leaderboards
Comprehensive rankings for AI models across all scientific categories
| Rank | Model | Score | Pass Rate | Evaluations | Trend |
|---|---|---|---|---|---|
1 | anthropic/claude-sonnet-4.5 | 87.1 | 100.0% | 22 | -27.3% |
2 | anthropic/claude-sonnet-4 | 85.8 | 95.0% | 20 | +79.6% |
3 | anthropic/claude-opus-4.1 | 86.4 | 95.5% | 22 | +217.7% |
4 | google/gemini-2.5-pro | 69.7 | 81.8% | 22 | +37.8% |
5 | openai/gpt-5 | 63.3 | 60.0% | 20 | +151.9% |
6 | x-ai/grok-code-fast-1 | 53.2 | 47.6% | 21 | -55.9% |
7 | x-ai/grok-4-fast | 39.3 | 28.0% | 25 | +12.4% |
8 | deepseek/deepseek-chat-v3.1 | 42.6 | 37.0% | 27 | +328.4% |
9 | openai/o3 | 33.3 | 13.6% | 22 | -3.2% |