HELM Classic - MMLU: leaderboard
The "MMLU" subtask of HELM Classic. A scenario from Stanford HELM Classic, the holistic evaluation suite scoring language models across accuracy, robustness and calibration on a broad set of tasks.
Metric: Exact Match (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 67 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Palmyra X (43B) | 60.91 | #1317 |
| 2 | LLaMA-65B | 58.37 | #1303 |
| 3 | Llama 2 70B | 58.17 | #1201 |
| 4 | Mistral-7B-v0.1 | 57.22 | #1326 |
| 5 | text-davinci-003 | 56.88 | #1146 |
| 6 | text-davinci-002 | 56.76 | #1225 |
| 7 | LLaMA-30B | 53.14 | #1423 |
| 8 | falcon-40B | 50.89 | #1442 |
| 9 | Llama 2 13B | 50.67 | #1436 |
| 10 | falcon-40B Instruct | 49.66 | #1458 |
| 11 | vicuna-13B-v1.3 | 46.16 | #1447 |
| 12 | mpt-30B Instruct | 44.44 | #1470 |
| 13 | mpt-30B | 43.67 | #1476 |
| 14 | vicuna-7B-v1.3 | 43.36 | #1507 |
| 15 | Llama 2 7B | 43.07 | #1520 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-classic-mmlu · How It Works · Data refreshed daily, snapshot 2026-10-20.