VMLU — leaderboard
Metric: Average (%). Source: vmlu.ai. 25 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | QwQ-32B | 76.13 |
| 2 | Llama 3 70B | 66.44 |
| 3 | GPT-4 | 65.53 |
| 4 | QwQ 32B-Preview | 63.9 |
| 5 | GPT-4o Mini | 62.87 |
| 6 | Gemma 2 9B (IT) | 59.04 |
| 7 | Qwen 2.5 7B Instruct | 57.51 |
| 8 | gemma-7B (IT) | 41.9 |
| 9 | Phi-3-small-128k-instruct | 41.24 |
| 10 | Phi-3-small-8k-instruct | 40.88 |
| 11 | Qwen-7B | 32.81 |
| 12 | Qwen 2 7B Instruct | 28.85 |
| 13 | gemma-2B (IT) | 27.72 |
| 14 | bloom-1b7 | 25.51 |
| 15 | bloom-7b1 | 25.41 |
Interactive version: theaggregate.ai/benchmark?slug=vmlu · How the rankings work · Data refreshed daily, snapshot 2026-07-22.