LA Leaderboard - CoQCat: leaderboard
Conversational question answering over Catalan texts, modelled on CoQA (CoQCat). From the LA Leaderboard, which evaluates models on Spanish, Catalan, Basque and Galician tasks with the lm-evaluation-harness; F1 (%), higher is better. Each model result file is a separate run.
Metric: F1 (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 69 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 2 9B | 78.51 | #924 |
| 2 | Llama 3.1 8B | 76.67 | #1130 |
| 3 | EuroLLM-9B | 76.1 | #1238 |
| 4 | Mistral-7B-v0.3 | 74.85 | #1321 |
| 5 | gemma-3-4B-pt Base | 74.37 | #1110 |
| 6 | salamandra-7B-instruct | 73.72 | #1392 |
| 7 | Qwen 2.5 7B | 73.47 | #895 |
| 8 | occiglot-7B-eu5 | 73.29 | |
| 9 | salamandra-7B | 71.94 | #1446 |
| 10 | Llama 3.2 3B | 70 | #1376 |
| 11 | Yi-1.5-9B | 69.93 | #1215 |
| 12 | SmolLM3-3B-Base | 66.85 | #1260 |
| 13 | Gemma 2 2B | 66.57 | #1466 |
| 14 | Gemma 2 9B (IT) | 66.11 | #795 |
| 15 | Qwen 2.5 3B | 65.8 | #1177 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=la-leaderboard-coqcat · How It Works · Data refreshed daily, snapshot 2026-11-05.