Polish Knowledge Benchmark - Military: leaderboard
Metric: Accuracy (%). Source: huggingface.co. 68 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 94.83 |
| 2 | Gemini 3.1 Flash Lite (Preview) | 92.67 |
| 3 | Gemini 3.6 Flash | 92.67 |
| 4 | Gemini 3.1 Flash Lite | 92.24 |
| 5 | Gemini 2.5 Flash | 91.38 |
| 6 | Gemini 3.5 Flash Lite | 90.95 |
| 7 | DeepSeek V4 Flash | 88.79 |
| 8 | Grok 4.1 Fast | 88.79 |
| 9 | Mistral Medium 3.1 | 88.36 |
| 10 | Gemma 4 31B (IT) | 86.64 |
| 11 | Qwen 3.5 397B A17B | 86.64 |
| 12 | Llama 4 Maverick | 86.21 |
| 13 | DeepSeek V3.2 | 84.05 |
| 14 | Llama 4 Scout | 84.05 |
| 15 | Gemini 2.0 Flash (001) | 83.19 |
Interactive version: theaggregate.ai/benchmark?slug=polish-knowledge-benchmark-military · How It Works · Data refreshed daily, snapshot 2026-09-19.