EuroEval Danish NLU - Multi Wiki QA DA: leaderboard
The "Multi Wiki QA DA" subtask of EuroEval Danish NLU. Danish-language EuroEval natural-language-understanding suite, separating NLU task performance from the broader all-task EuroEval aggregate.
Metric: Character F1 (%). Source: euroeval.com. Saturation forecast: Estimated already saturated. 270 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Mistral Small 3.2 | 82.73 | #605 |
| 2 | Magistral Small 1.2 | 82.72 | #570 |
| 3 | SOLAR-10.7B-v1.0 | 82.34 | #1253 |
| 4 | Qwen 3 14B (Non-reasoning) | 82.01 | #537 (Qwen 3 14B) |
| 5 | gemma-3-27B-pt | 81.81 | #502 |
| 6 | Qwen 3 32B (Non-reasoning) | 81.7 | #443 (Qwen 3 32B) |
| 7 | Bielik-11B-v2.3-Instruct | 81.65 | #693 |
| 8 | Voxtral-Small-24B-2507 | 80.98 | #581 |
| 9 | Ministral-3-14B-Reasoning-2512 | 80.83 | #540 |
| 10 | Llama 3.1 70B | 80.7 | #592 |
| 11 | Mistral Small 3.1 | 80.39 | #610 |
| 12 | Olmo-3.1-32B-Instruct-SFT | 80.11 | #879 |
| 13 | Qwen 2.5 72B Instruct | 79.64 | #454 |
| 14 | Olmo-3.1-32B-Instruct-DPO | 79.52 | #845 |
| 15 | Llama 3.1 8B | 78.62 | #1151 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-danish-nlu-multi-wiki-qa-da · How It Works · Data refreshed daily, snapshot 2026-10-20.