HELM MMLU - Global Facts: leaderboard
The "Global Facts" 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) | 80 | #291 |
| 2 | Gemini 1.5 Pro (002) | 77 | #350 |
| 3 | Claude 3.5 Sonnet (20240620) | 72 | #298 |
| 4 | Llama 3.1 405B Instruct | 71 | #461 |
| 5 | GPT-4o (2024-08-06) | 69 | #335 |
| 6 | DeepSeek V3 | 68 | #323 |
| 7 | Gemini 2.0 Flash (Preview) | 66 | #366 |
| 8 | Claude 3 Opus (20240229) | 66 | #413 |
| 9 | Gemini 1.5 Pro (001) | 66 | #446 |
| 10 | GPT-4o (2024-05-13) | 64 | #326 |
| 11 | GPT-4 (0613) | 62 | #494 |
| 12 | Llama 3.1 70B Instruct | 61 | #566 |
| 13 | Qwen 2.5 72B Instruct | 61 | #458 |
| 14 | Llama 3.2 90B Vision Instruct | 60 | #626 |
| 15 | Llama 3.3 70B Instruct | 58 | #536 |
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-global-facts · How It Works · Data refreshed daily, snapshot 2026-10-23.