HELM MMLU - Moral Disputes: leaderboard
Metric: EM. Source: crfm.stanford.edu. 79 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude 3.5 Sonnet (20240620) | 88.73 |
| 2 | GPT-4o (2024-05-13) | 88.15 |
| 3 | Claude 3 Opus (20240229) | 88.15 |
| 4 | DeepSeek V3 | 87.28 |
| 5 | Claude 3.5 Sonnet (20241022) | 86.99 |
| 6 | GPT-4o (2024-08-06) | 86.99 |
| 7 | Llama 3.1 405B Instruct | 86.99 |
| 8 | GPT-4 (0613) | 86.71 |
| 9 | Gemini 2.0 Flash (Preview) | 86.42 |
| 10 | GPT-4 Turbo | 86.13 |
| 11 | GPT-4 Turbo (Preview) | 85.84 |
| 12 | Gemini 1.5 Pro (002) | 85.55 |
| 13 | Llama 3.3 70B Instruct | 85.26 |
| 14 | Qwen 2 72B Instruct | 85.26 |
| 15 | Llama 3.2 90B Vision Instruct | 84.97 |
Interactive version: theaggregate.ai/benchmark?slug=helm-mmlu-moral-disputes · How It Works · Data refreshed daily, snapshot 2026-09-08.