BenCzechMark - NER: leaderboard
BenCzechMark NER category of Czech tasks. Scores are duel win scores (0-100): the share of pairwise statistical duels a model wins against every other model on the category's tasks, averaged over tasks, so they shift as the roster changes; higher is better.
Metric: Duel Win Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 79 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Muse Glimmer 30B | 98.08 | #191 |
| 2 | Llama 3.1 405B Instruct | 94.23 | #470 |
| 3 | Llama 3.1 70B | 93.59 | #588 |
| 4 | Qwen 2.5 72B | 92.31 | #512 |
| 5 | Llama 3.3 70B Instruct | 87.18 | #541 |
| 6 | Llama 4 Scout Instruct | 87.18 | #547 |
| 7 | DeepSeek R1 0528 | 86.54 | #232 |
| 8 | DeepSeek V4 Flash (0731) | 86.54 | #84 |
| 9 | DeepSeek V3 (0324) | 85.9 | #349 |
| 10 | c4ai-command-a-03-2025 | 85.26 | #468 |
| 11 | Llama 3.1 70B Instruct | 83.97 | #567 |
| 12 | Qwen 3.6 27B | 80.77 | #182 |
| 13 | Qwen 2.5 7B | 77.56 | #895 |
| 14 | c4ai-command-r-08-2024 | 77.56 | #949 |
| 15 | Mistral Large 2 (Nov) Instruct (2411) | 75 | #480 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=benczechmark-ner · How It Works · Data refreshed daily, snapshot 2026-11-05.