HELM AFR - MMLU Clinical Knowledge (Tsonga): leaderboard

Metric: EM. Source: crfm.stanford.edu. 22 models tracked.

Top models

#ModelScore
1Gemini 2.0 Flash70.19
2Claude 3.7 Sonnet (20250219)65.28
3GPT-4o (2024-08-06)64.15
4DeepSeek V361.13
5Gemini 2.0 Flash Lite61.13
6GPT-4 Turbo55.47
7Llama 3.1 405B Instruct54.34
8Llama 3.1 70B Instruct51.32
9Llama 3.3 70B Instruct48.3
10Gemma 2 27B (IT)43.77
11Qwen 2.5 72B Instruct42.64
12Claude 3.5 Haiku (20241022)42.26
13Llama 3.1 8B Instruct38.87
14GPT-3.5 Turbo (0125)35.85
15Gemma 2 9B (IT)34.72

Interactive version: theaggregate.ai/benchmark?slug=helm-afr-mmlu-clinical-knowledge-tsonga · How It Works · Data refreshed daily, snapshot 2026-09-08.