Open PL LLM - RAG — leaderboard

Metric: Average RAG Score (%). Source: huggingface.co. 287 models tracked.

Top models

#ModelScore
1Bielik-11B-v2.2-Instruct76.92
2Bielik-11B-v2.1-Instruct76.79
3Bielik-11B-v2.3-Instruct76.2
4Mistral Large 2 (Jul)75.78
5MSH-v1-Bielik-v2.3-Instruct-MedIT-merge75.62
6Bielik-11B-v2.0-Instruct75.5
7Qwen 2 72B75.48
8Mistral Large 2 (Nov) Instruct (2411)74.96
9Llama 3.3 70B Instruct74.62
10Llama 3.1 405B Instruct FP874.34
11Gemma 3 27B (IT)73.92
12Llama 4 Scout Instruct73.89
13QwQ 32B-Preview73.8
14Qwen 3 14B73.73
15Llama 3.1 70B Instruct73.67

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