EuroEval Dutch NLU - ScaLA NL: leaderboard
The "ScaLA 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: Matthews correlation coefficient (%). Source: euroeval.com. Saturation forecast: Estimated already saturated. 332 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Flash Lite (Preview) | 77.57 | #194 |
| 2 | Gemini 3 Pro (Preview) | 77.07 | #66 |
| 3 | Gemini 3 Flash (Preview) (Non-reasoning) | 75.69 | #82 (Gemini 3 Flash (Preview)) |
| 4 | Claude 3.7 Sonnet (20250219) (Thinking) | 74.66 | #201 (Claude 3.7 Sonnet (20250219)) |
| 5 | Gemini 2.5 Pro | 73.47 | #152 |
| 6 | Gemini 3 Flash (Preview) (Thinking) | 73.11 | #82 (Gemini 3 Flash (Preview)) |
| 7 | Llama 3.1 405B | 72.93 | #466 |
| 8 | Gemini 2.5 Pro (Preview 03-25) | 70.76 | #123 |
| 9 | Claude Sonnet 4.5 (Thinking) | 70.03 | #142 (Claude Sonnet 4.5) |
| 10 | GPT-5.4 Mini (2026-03-17) (High) | 69.61 | #169 (GPT-5.4 Mini (2026-03-17)) |
| 11 | Llama 3.1 405B Instruct FP8 | 69.11 | #389 |
| 12 | Gemma 4 31B (IT) | 69.01 | #176 |
| 13 | O3 (2025-04-16) | 68.86 | #121 |
| 14 | Qwen 3 235B A22B 2507 Instruct | 68.5 | #300 |
| 15 | Gemini 2.5 Flash (Preview 04-17) | 68.47 | #237 |
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-scala-nl · How It Works · Data refreshed daily, snapshot 2026-10-20.