BenCzechMark - Sentiment: leaderboard
BenCzechMark Sentiment 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: Around February 2027. 79 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | c4ai-command-a-03-2025 | 86.86 | #468 |
| 2 | Gemma 3 27B (IT) | 84.94 | #528 |
| 3 | c4ai-command-r-08-2024 | 81.73 | #949 |
| 4 | Mistral Large 2 (Nov) Instruct (2411) | 79.81 | #480 |
| 5 | aya-expanse-32B | 77.88 | #687 |
| 6 | DeepSeek V3 (0324) | 76.92 | #349 |
| 7 | Gemma 2 27B (IT) | 76.28 | #652 |
| 8 | Llama 3.3 70B Instruct | 75.32 | #541 |
| 9 | Llama 3.1 70B Instruct | 75 | #567 |
| 10 | Qwen 2.5 72B Instruct | 74.68 | #469 |
| 11 | Qwen 2.5 72B | 74.68 | #512 |
| 12 | Mixtral 8x22B Instruct | 73.4 | #738 |
| 13 | MiniMax M1 80k | 73.08 | #437 |
| 14 | aya-expanse-8B | 72.44 | #914 |
| 15 | Qwen 2.5 32B Instruct | 72.12 | #509 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=benczechmark-sentiment · How It Works · Data refreshed daily, snapshot 2026-11-05.