OpenCompass Reasoning - Humanities - English (CompassBench 2407): leaderboard

Metric: Score (%). Source: rank.opencompass.org.cn. 36 models tracked.

Top models

#ModelScore
1Claude 3.5 Sonnet (20240620)60.2
2GPT-4o (2024-05-13)59
3Mistral Large 2 (Jul)55.5
4Mistral Large51.9
5GPT-4o Mini (2024-07-18)51.3
6Yi Large50.2
7GLM-4 (0520)49.9
8Llama 3.1 70B Instruct49.7
9Gemma 2 27B (IT)49.1
10Yi 1.5 34B Chat48.2
11DeepSeek V2 Chat (0628)47.1
12Moonshot v1 8K46.4
13GLM-4 9B Chat45.8
14Yi-1.5-6B-Chat44.3
15Llama 3.1 8B Instruct43.9

Interactive version: theaggregate.ai/benchmark?slug=opencompass-reasoning-humanities-english-compassbench-2407 · How It Works · Data refreshed daily, snapshot 2026-09-19.