HELM MMLU - High School European History: leaderboard

Metric: EM. Source: crfm.stanford.edu. 79 models tracked.

Top models

#ModelScore
1GPT-4o (2024-08-06)91.52
2GPT-4o (2024-05-13)90.3
3Claude 3.5 Sonnet (20241022)89.7
4Claude 3.5 Sonnet (20240620)89.7
5GPT-4 Turbo (Preview)89.7
6Gemini 1.5 Pro (002)88.48
7Qwen 2.5 72B Instruct88.48
8GPT-4o Mini (2024-07-18)88.48
9Llama 3.1 405B Instruct88.48
10Claude 3 Opus (20240229)88.48
11Qwen 1.5 110B88.48
12DeepSeek V387.88
13GPT-4 (0613)87.88
14Yi Large (Preview)87.88
15Llama 3.1 70B Instruct87.27

Interactive version: theaggregate.ai/benchmark?slug=helm-mmlu-high-school-european-history · How It Works · Data refreshed daily, snapshot 2026-09-08.