Polish Knowledge Benchmark - Biographies: leaderboard
Topic category of the Polish Knowledge Benchmark (PolishQA) covering biographies of notable Poles. The benchmark has 15,000 synthetic four-option questions about Poland generated by DeepSeek 3.2 and difficulty-rated by Qwen3.5-397B-A17B. Accuracy is correct over total in the category, 0-100, higher is better.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 69 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 95.53 | #90 |
| 2 | Mistral Large 3 | 90.12 | #410 |
| 3 | Gemini 3.6 Flash | 89.65 | #57 |
| 4 | Gemini 3.1 Flash Lite (Preview) | 88.94 | #201 |
| 5 | Gemini 3.1 Flash Lite | 87.71 | #200 |
| 6 | Qwen 3.5 397B A17B | 87.29 | #154 |
| 7 | DeepSeek V4 Flash | 86.59 | #155 |
| 8 | DeepSeek V3.2 | 85.65 | #210 |
| 9 | Mistral Medium 3.1 | 84.47 | #418 |
| 10 | DeepSeek V3 (0324) | 82.35 | #349 |
| 11 | Mistral Large 4 | 82.12 | #106 |
| 12 | Gemini 2.5 Flash | 81.88 | #255 |
| 13 | Gemini 2.0 Flash (001) | 81.18 | #365 |
| 14 | Gemini 3.5 Flash Lite | 80.24 | #174 |
| 15 | Grok 4.1 Fast | 80 | #227 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=polish-knowledge-benchmark-biographies · How It Works · Data refreshed daily, snapshot 2026-11-05.