Open PL LLM - Generative: leaderboard
Open PL LLM aggregate over generative Polish tasks.
Metric: Average Generative Score (%). Source: huggingface.co. Saturation forecast: Around November 2026. 287 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Mistral Large 2 (Nov) Instruct (2411) | 73.07 | #465 |
| 2 | Llama 3.1 405B Instruct FP8 | 72.51 | #389 |
| 3 | Mistral Large 2 (Jul) | 72.21 | #484 |
| 4 | Llama 3.3 70B Instruct | 70.23 | #534 |
| 5 | Llama 3.1 70B Instruct | 69.99 | #561 |
| 6 | Qwen 2.5 72B Instruct | 69.79 | #454 |
| 7 | Qwen 2.5 72B | 69.48 | #499 |
| 8 | Qwen 2 72B Instruct | 69.22 | #576 |
| 9 | Gemma 3 27B (IT) | 69.13 | #523 |
| 10 | Gemma 2 27B (IT) | 69.13 | #645 |
| 11 | Llama 3.1 Nemotron 70B Instruct | 69.13 | #554 |
| 12 | Mixtral 8x22B Instruct (v0.1) | 68.64 | #756 |
| 13 | QwQ 32B-Preview | 68.35 | #651 |
| 14 | Qwen 2 72B | 68.3 | #521 |
| 15 | Qwen 2.5 32B | 68.28 | #546 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-pl-llm-generative · How It Works · Data refreshed daily, snapshot 2026-10-20.