Polish Knowledge Benchmark - Theatre: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 68 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 91.54 |
| 2 | Gemini 3.6 Flash | 90 |
| 3 | Gemini 3.1 Flash Lite | 87.69 |
| 4 | Gemini 3.1 Flash Lite (Preview) | 86.92 |
| 5 | Gemini 3.5 Flash Lite | 86.54 |
| 6 | Gemini 2.5 Flash | 81.15 |
| 7 | Mistral Large 3 | 81.15 |
| 8 | DeepSeek V4 Flash | 81.15 |
| 9 | DeepSeek V3.2 | 80 |
| 10 | Gemma 4 31B (IT) | 79.23 |
| 11 | Qwen 3.5 397B A17B | 79.23 |
| 12 | Mistral Medium 3.1 | 78.85 |
| 13 | Gemini 2.0 Flash (001) | 77.69 |
| 14 | Grok 4.1 Fast | 76.54 |
| 15 | DeepSeek V3 (0324) | 76.15 |
Interactive version: theaggregate.ai/benchmark?slug=polish-knowledge-benchmark-theatre · How It Works · Data refreshed daily, snapshot 2026-09-19.