LA Leaderboard - ARC Catalan: leaderboard
Catalan translation of the ARC grade-school science multiple-choice questions (CatalanBench). 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 | 67.87 | #924 |
| 2 | Gemma 2 9B (IT) | 64.52 | #795 |
| 3 | EuroLLM-9B-Instruct | 63.9 | #1047 |
| 4 | EuroLLM-9B | 63.61 | #1238 |
| 5 | Qwen3-8B-Base | 60.99 | #626 |
| 6 | Mistral 7B Instruct (v0.3) | 60.94 | #1255 |
| 7 | salamandra-7B | 60.48 | #1446 |
| 8 | Llama 3.1 8B Instruct | 59.78 | #1017 |
| 9 | gemma-3-4B-pt Base | 59.78 | #1110 |
| 10 | Mistral-7B-v0.3 | 59.7 | #1321 |
| 11 | salamandra-7B-instruct | 58.99 | #1392 |
| 12 | Gemma 3 4B (IT) | 58.37 | #997 |
| 13 | Llama 3.1 8B | 58.34 | #1130 |
| 14 | occiglot-7B-eu5 | 56.43 | |
| 15 | occiglot-7B-eu5-instruct | 54.79 |
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-arc-catalan · How It Works · Data refreshed daily, snapshot 2026-11-05.