Open PL LLM - PolQA Open Book (generative, 5-shot): leaderboard

Metric: Levenshtein Similarity (%). Source: huggingface.co. 318 models tracked.

Top models

#ModelScore
1MSH-v1-Bielik-v2.3-Instruct-MedIT-merge92.65
2Bielik-11B-v2.1-Instruct92.39
3Mixtral-8x22B-v0.192.21
4Bielik-11B-v2.3-Instruct92.19
5Bielik-11B-v2.2-Instruct92.12
6Llama 3 70B92.09
7Llama 4 Scout Base91.92
8Bielik-11B-v2.0-Instruct91.92
9Llama 3.1 70B91.68
10Bielik-11B-v291.46
11Mistral Large 2 (Nov) Instruct (2411)91.45
12Llama 3.1 405B Instruct FP890.99
13Mistral-Small-24B-Base-250190.71
14Mistral Large 2 (Jul)90.57
15Bielik-11B-v2.5-Instruct90.5

Interactive version: theaggregate.ai/benchmark?slug=open-pl-llm-polqa-open-book-generative-5-shot · How It Works · Data refreshed daily, snapshot 2026-09-19.