LA Leaderboard - ClinTreatES: leaderboard
Spanish clinical cases for which the model writes the treatment (ClinTreatES). From the LA Leaderboard, which evaluates models on Spanish, Catalan, Basque and Galician tasks with the lm-evaluation-harness; semantic answer similarity, the cosine similarity of sentence embeddings x100, higher is better. Each model result file is a separate run.
Metric: Semantic Answer Similarity (x100). Source: huggingface.co. Saturation forecast: Estimated already saturated. 68 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 7B Instruct | 67.46 | #863 |
| 2 | Gemma 3 4B (IT) | 65.6 | #997 |
| 3 | Qwen 2.5 1.5B Instruct | 65.27 | #1393 |
| 4 | Mistral 7B Instruct (v0.3) | 65.24 | #1255 |
| 5 | Yi-1.5-9B Chat | 64.88 | #1032 |
| 6 | salamandra-7B-instruct | 64.07 | #1392 |
| 7 | occiglot-7B-eu5-instruct | 63.93 | |
| 8 | Qwen 2.5 7B | 63.8 | #895 |
| 9 | Gemma 3 1B (IT) | 63.7 | #1481 |
| 10 | Gemma 2 9B (IT) | 62.83 | #795 |
| 11 | Llama 3.1 8B | 62.41 | #1130 |
| 12 | Qwen 2.5 3B | 61.88 | #1177 |
| 13 | Gemma 2 9B | 61.87 | #924 |
| 14 | Teuken-7B-instruct-commercial-v0.4 | 61.68 | #1421 |
| 15 | SmolLM2-1.7B-Instruct | 60.81 | #1541 |
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-clintreates · How It Works · Data refreshed daily, snapshot 2026-11-05.