Open PL LLM - PolQA Closed Book (generative, 5-shot): leaderboard
Answering Polish PolQA questions from memory, without passages. Open PL LLM Leaderboard (SpeakLeash), 5-shot, generative variant: the answer is extracted from generated text; Levenshtein similarity to the reference answer (%), higher is better. Runs not available on the Hugging Face Hub are excluded.
Metric: Levenshtein Similarity (%). Source: huggingface.co. Saturation forecast: Around February 2027. 319 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 3.1 405B Instruct FP8 | 85.98 | #400 |
| 2 | Mistral Large 2 (Jul) | 85.05 | #490 |
| 3 | Mistral Large 2 (Nov) Instruct (2411) | 85.05 | #480 |
| 4 | Bielik-11B-v2 | 85.05 | #1309 |
| 5 | Bielik-11B-v2.3-Instruct | 84.11 | #700 |
| 6 | Bielik-11B-v2.2-Instruct | 83.59 | #669 |
| 7 | Llama 3.1 70B | 83.49 | #588 |
| 8 | Llama 4 Scout Base | 83.39 | #708 |
| 9 | Bielik-11B-v2.1-Instruct | 83.28 | #679 |
| 10 | Llama 3 70B | 83.07 | #632 |
| 11 | Bielik-11B-v2.5-Instruct | 82.97 | #575 |
| 12 | Bielik-11B-v2.0-Instruct | 82.45 | #750 |
| 13 | Mixtral 8x22B | 81.62 | #725 |
| 14 | Llama 3.3 70B Instruct | 80.37 | #541 |
| 15 | Llama 3.1 Nemotron 70B Instruct | 80.17 | #563 |
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-polqa-closed-book-generative-5-shot · How It Works · Data refreshed daily, snapshot 2026-11-05.