MT-Bench PL - Overall: leaderboard
Metric: Judge Score (0-10). Source: huggingface.co. 50 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemma 3 27B (IT) | 9.28 |
| 2 | Mistral Small 3.1 | 9.18 |
| 3 | Phi-4 | 9.07 |
| 4 | Gemma 3 12B (IT) | 8.97 |
| 5 | Qwen 2.5 32B Instruct | 8.86 |
| 6 | Qwen 2 72B Instruct | 8.78 |
| 7 | Mistral Small 3 | 8.72 |
| 8 | Mistral Large 2 (Jul) | 8.66 |
| 9 | Gemma 2 27B (IT) | 8.62 |
| 10 | aya-expanse-32B | 8.62 |
| 11 | Mistral-Small-Instruct-2409 | 8.56 |
| 12 | Qwen 2.5 14B Instruct | 8.33 |
| 13 | Mixtral 8x22B | 8.23 |
| 14 | Gemma 3 4B (IT) | 8.22 |
| 15 | Llama 3.1 405B Instruct | 8.17 |
Interactive version: theaggregate.ai/benchmark?slug=mt-bench-pl-overall · How It Works · Data refreshed daily, snapshot 2026-09-05.