EuroEval Dutch NLU - SQuAD NL: leaderboard
The "SQuAD NL" subtask of EuroEval Dutch NLU. Dutch-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. 325 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Ministral-3-14B-Reasoning-2512 | 84.22 | #540 |
| 2 | Qwen3.5-4B-Base | 84.17 | #798 |
| 3 | Llama 3.1 405B | 84.07 | #466 |
| 4 | Llama 3.1 70B | 82.52 | #592 |
| 5 | Gemma 2 27B | 82.44 | #806 |
| 6 | Qwen 3 32B (Non-reasoning) | 82.41 | #443 (Qwen 3 32B) |
| 7 | Magistral Small 1.2 | 82.21 | #570 |
| 8 | gemma-3-27B-pt | 81.67 | #502 |
| 9 | Yi 1.5 34B | 81.65 | #807 |
| 10 | Bielik-11B-v2.3-Instruct | 81.43 | #693 |
| 11 | Mistral Small 3.1 | 81.17 | #610 |
| 12 | Mistral Small 3.2 | 81.05 | #605 |
| 13 | Qwen 3 14B (Non-reasoning) | 80.42 | #537 (Qwen 3 14B) |
| 14 | Llama 2 70B Base | 80.42 | #1025 |
| 15 | EuroLLM-9B | 79.92 | #1254 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-dutch-nlu-squad-nl · How It Works · Data refreshed daily, snapshot 2026-10-20.