OpenCompass Knowledge - Humanities - English (CompassBench 2409): leaderboard

Metric: Score (%). Source: rank.opencompass.org.cn. 30 models tracked.

Top models

#ModelScore
1GPT-4o (2024-08-06)92
2GPT-4o (2024-05-13)91.2
3Step 2 16K89.6
4Claude 3.5 Sonnet (20240620)88
5Llama 3.1 70B Instruct85.6
6Qwen 2.5 72B Instruct85.6
7Mistral Large 2 (Jul)85.6
8GLM-4 Plus85.6
9Llama 3.1 405B Instruct FP884.8
10GPT-4o Mini (2024-07-18)83.2
11Yi Large76
12DeepSeek V2.576
13Mistral-Small-Instruct-240972.8
14Gemma 2 9B (IT)69.6
15Llama 3.1 8B Instruct66.4

Interactive version: theaggregate.ai/benchmark?slug=opencompass-knowledge-humanities-english-compassbench-2409 · How It Works · Data refreshed daily, snapshot 2026-09-19.