HELM MMLU - Overall: leaderboard
The "Overall" subtask of HELM MMLU. HELM Classic MMLU group leaderboard from the public HELM release artifacts.
Metric: EM. Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 79 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude 3.5 Sonnet (20241022) | 87.27 | #291 |
| 2 | DeepSeek V3 | 87.2 | #323 |
| 3 | Gemini 1.5 Pro (002) | 86.88 | #350 |
| 4 | Claude 3.5 Sonnet (20240620) | 86.54 | #298 |
| 5 | Claude 3 Opus (20240229) | 84.57 | #413 |
| 6 | Llama 3.1 405B Instruct | 84.53 | #461 |
| 7 | GPT-4o (2024-08-06) | 84.33 | #335 |
| 8 | GPT-4o (2024-05-13) | 84.17 | #326 |
| 9 | Qwen 2.5 72B Instruct | 83.37 | #458 |
| 10 | Gemini 1.5 Pro (001) | 82.71 | #446 |
| 11 | GPT-4 (0613) | 82.44 | #494 |
| 12 | Qwen 2 72B Instruct | 82.35 | #579 |
| 13 | Nova Pro | 82 | #573 |
| 14 | GPT-4 Turbo | 81.27 | #406 |
| 15 | Llama 3.2 90B Vision Instruct | 80.27 | #626 |
No result here: #4 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-mmlu-overall · How It Works · Data refreshed daily, snapshot 2026-10-23.