LA Leaderboard - Belebele Basque: leaderboard
Belebele four-option reading comprehension on FLORES-200 passages, Basque version. From the LA Leaderboard, which evaluates models on Spanish, Catalan, Basque and Galician tasks with the lm-evaluation-harness; accuracy (%), higher is better. Each model result file is a separate run.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 69 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 2 9B (IT) | 81.78 | #795 |
| 2 | Gemma 2 9B | 79.44 | #924 |
| 3 | Llama 3.1 8B Instruct | 72.89 | #1017 |
| 4 | Gemma 3 4B (IT) | 63.89 | #997 |
| 5 | gemma-3-4B-pt Base | 62.44 | #1110 |
| 6 | Llama 3.1 8B | 60.67 | #1130 |
| 7 | salamandra-7B-instruct | 57.56 | #1392 |
| 8 | Qwen 2.5 7B | 54.89 | #895 |
| 9 | Qwen 2.5 7B Instruct | 52.89 | #863 |
| 10 | Llama 3.2 3B Instruct | 50.44 | #1327 |
| 11 | aya-expanse-8B | 47.89 | #914 |
| 12 | Qwen 2.5 3B Instruct | 46.33 | #1159 |
| 13 | Gemma 2 2B (IT) | 45.89 | #1328 |
| 14 | Qwen 2.5 3B | 42 | #1177 |
| 15 | Llama 3.2 3B | 41.78 | #1376 |
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-belebele-basque · How It Works · Data refreshed daily, snapshot 2026-11-05.