EuroEval Catalan NLU - Guia CAT: leaderboard
The "Guia CAT" subtask of EuroEval Catalan NLU. Catalan-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. 212 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3.5 9B | 77.18 | #379 |
| 2 | GPT-5.4 Mini (2026-03-17) (High) | 76.13 | #169 (GPT-5.4 Mini (2026-03-17)) |
| 3 | GLM-4.7 Flash | 76.04 | #509 |
| 4 | Tiny Aya Global | 75.19 | #1325 |
| 5 | tiny-aya-water | 74.43 | |
| 6 | Gemini 3 Flash (Preview) (Thinking) | 74.01 | #82 (Gemini 3 Flash (Preview)) |
| 7 | GPT-5 Mini (2025-08-07) | 73.82 | #172 |
| 8 | Bielik-11B-v2.3-Instruct | 73.6 | #693 |
| 9 | Qwen 3.5 27B FP8 | 73.14 | #164 |
| 10 | Qwen 3.5 35B A3B FP8 | 72.81 | #362 |
| 11 | Gemma 4 31B (IT) | 72.75 | #176 |
| 12 | Qwen 3.5 4B | 72.49 | #475 |
| 13 | GPT-4.1 | 72.36 | #248 |
| 14 | Claude Sonnet 4.6 | 72.21 | #90 |
| 15 | Ministral-3-14B-Reasoning-2512 | 71.85 | #540 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-catalan-nlu-guia-cat · How It Works · Data refreshed daily, snapshot 2026-10-20.