MT-Bench PL - Coding — leaderboard

Metric: Judge Score (0-10). Source: huggingface.co. 50 models tracked.

Top models

#ModelScore
1Mistral Small 3.18.3
2Gemma 3 12B (IT)8.25
3Gemma 3 27B (IT)8.1
4Qwen 2.5 32B Instruct7.95
5Mistral Small 37.95
6Qwen 2 72B Instruct7.8
7Phi-47.6
8Gemma 2 27B (IT)7.45
9Llama 3.1 405B Instruct7.25
10Mistral-Small-Instruct-24097.1
11Mistral Large 2 (Jul)6.75
12Qwen 2.5 14B Instruct6.7
13Mixtral 8x22B6.45
14Llama 3.1 70B Instruct6.25
15Bielik-11B-v2.3-Instruct6.25

Interactive version: theaggregate.ai/benchmark?slug=mt-bench-pl-coding · How the rankings work · Data refreshed daily, snapshot 2026-07-22.