MT-Bench PL - Extraction — leaderboard
Metric: Judge Score (0-10). Source: huggingface.co. 50 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemma 3 27B (IT) | 9.9 |
| 2 | Qwen 2.5 32B Instruct | 9.9 |
| 3 | Mistral Small 3 | 9.9 |
| 4 | Mistral Large 2 (Jul) | 9.9 |
| 5 | Llama 3.1 70B Instruct | 9.85 |
| 6 | Llama 3.1 405B Instruct | 9.85 |
| 7 | Mistral Small 3.1 | 9.8 |
| 8 | Qwen 2 72B Instruct | 9.8 |
| 9 | Gemma 2 27B (IT) | 9.6 |
| 10 | Gemma 3 12B (IT) | 9.55 |
| 11 | Mixtral 8x22B | 9.55 |
| 12 | Bielik-11B-v2.3-Instruct | 9.43 |
| 13 | Phi-4 | 9.3 |
| 14 | Bielik-11B-v2.2-Instruct | 9.3 |
| 15 | Qwen 2.5 14B Instruct | 9.25 |
Interactive version: theaggregate.ai/benchmark?slug=mt-bench-pl-extraction · How the rankings work · Data refreshed daily, snapshot 2026-07-22.