PhysicsFinals — leaderboard

PhysicsFinals evaluates model capability on science tasks from the linked upstream source with Score as the primary reported metric.

Metric: Score (self-reported). Source: benchmarklist.com. Status: saturation imminent. 33 models tracked.

Top models

#ModelScore
1O3 Mini59.9
2O1 Mini43.6
3Gemini 1.5 Pro38.4
4GPT-4o36.7
5Claude 3.5 Sonnet34.7
6Llama 3.3 70B Instruct31.5
7Phi-429.1
8Qwen 2.5 72B Instruct28.7
9Qwen 2.5 32B27.6
10Qwen 2.5 7B20.4
11Qwen 2.5 14B19.6
12Gemma 2 27B18.3
13Yi 1.5 34B17.4
14QwQ 32B-Preview12.1
15Gemma 2 9B11.9

Interactive version: theaggregate.ai/benchmark?slug=physicsfinals · How the rankings work · Data refreshed daily, snapshot 2026-07-22.