Polish Legal RAG - Source Classification: leaderboard
Metric: Classification score (0-100). Source: huggingface.co. 37 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 Mini | 96.71 |
| 2 | GPT-4.1 | 96.39 |
| 3 | Qwen 3 235B A22B 2507 Instruct | 96.21 |
| 4 | GPT-OSS-120B | 96.17 |
| 5 | Bielik-11B-v2.5-Instruct | 96.12 |
| 6 | Llama 4 Maverick Instruct | 96.08 |
| 7 | GPT-4o | 96.05 |
| 8 | DeepSeek V3 (0324) | 96.03 |
| 9 | Kimi K2 | 95.91 |
| 10 | Llama 3.3 70B Instruct | 95.89 |
| 11 | Claude Sonnet 4 | 95.88 |
| 12 | Llama 4 Scout Instruct | 95.86 |
| 13 | Qwen 3 32B | 95.86 |
| 14 | GPT-4o Mini | 95.8 |
| 15 | GPT-OSS-20B | 95.78 |
Interactive version: theaggregate.ai/benchmark?slug=polish-legal-rag-source-classification · How It Works · Data refreshed daily, snapshot 2026-09-19.