HELM AFR - MMLU Virology: leaderboard
Metric: EM. Source: crfm.stanford.edu. 22 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 2.0 Flash | 52.35 |
| 2 | Claude 3.7 Sonnet (20250219) | 52.08 |
| 3 | GPT-4o (2024-08-06) | 47.21 |
| 4 | DeepSeek V3 | 47.15 |
| 5 | Gemini 2.0 Flash Lite | 47.1 |
| 6 | Gemma 2 27B (IT) | 40.42 |
| 7 | Claude 3.5 Haiku (20241022) | 38.88 |
| 8 | Llama 3.1 70B Instruct | 38.72 |
| 9 | Llama 3.3 70B Instruct | 38.12 |
| 10 | GPT-4 Turbo | 37.4 |
| 11 | Gemma 2 9B (IT) | 37.02 |
| 12 | Qwen 2.5 72B Instruct | 35.32 |
| 13 | GPT-3.5 Turbo (0125) | 30.67 |
| 14 | Llama 3.1 405B Instruct | 30.56 |
| 15 | Mistral 7B Instruct (v0.3) | 30.56 |
Interactive version: theaggregate.ai/benchmark?slug=helm-afr-mmlu-virology · How It Works · Data refreshed daily, snapshot 2026-09-08.