LA Leaderboard - XNLI Catalan: leaderboard
XNLI natural language inference in Catalan. 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 | Llama 3.1 8B Instruct | 52.49 | #1017 |
| 2 | occiglot-7B-eu5-instruct | 51.85 | |
| 3 | Mistral 7B Instruct (v0.3) | 51.2 | #1255 |
| 4 | occiglot-7B-eu5 | 50.68 | |
| 5 | Qwen3-8B-Base | 50.64 | #626 |
| 6 | Teuken-7B-Instruct-v0.4 | 49.4 | #1408 |
| 7 | salamandra-7B | 49.24 | #1446 |
| 8 | salamandra-7B-instruct | 49.04 | #1392 |
| 9 | Llama 3.1 8B | 48.92 | #1130 |
| 10 | Gemma 2 2B | 48.84 | #1466 |
| 11 | Qwen 2.5 7B | 48.8 | #895 |
| 12 | Yi-1.5-9B | 48.76 | #1215 |
| 13 | SmolLM3-3B | 48.76 | #1239 |
| 14 | Gemma 2 9B (IT) | 48.51 | #795 |
| 15 | Teuken-7B-instruct-commercial-v0.4 | 48.51 | #1421 |
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-xnli-catalan · How It Works · Data refreshed daily, snapshot 2026-11-05.