PHYBench — leaderboard
Metric: EED Score. Source: www.phybench.cn. 32 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.5 Pro | 56.54 |
| 2 | Grok 4 | 53.03 |
| 3 | GPT-5 | 51.07 |
| 4 | Gemini 2.5 Pro (03-25) | 49.46 |
| 5 | O3 (High) | 46.37 |
| 6 | Gemini 2.5 Flash | 44.84 |
| 7 | DeepSeek R1 | 44.25 |
| 8 | O4 Mini (High) | 41.94 |
| 9 | GLM-4.5 | 37.86 |
| 10 | O3 Mini (High) | 37.26 |
| 11 | O4 Mini | 36.44 |
| 12 | Qwen 3 Max | 33.64 |
| 13 | Kimi K2 | 33.56 |
| 14 | O3 Mini | 33.26 |
| 15 | Grok 3 | 32 |
Interactive version: aibenchmarks.dev/benchmark?slug=phybench · How the rankings work · Data refreshed daily, snapshot 2026-07-22.