LatamBoard - Spanish WNLI: leaderboard
The "Spanish WNLI" subtask of LatamBoard. Leaderboard for model performance on Latin American and Iberian language tasks, including Spanish and Portuguese understanding, TELEIA Spanish exams, FLORES/OPUS translation, and structured or image extraction.
Metric: Score (%). Source: huggingface.co. Saturation forecast: Around January 2027. 35 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Phi-3-medium-4k-instruct | 81.69 | #952 |
| 2 | gemma-3-12B-pt | 80.28 | #686 |
| 3 | SOLAR-10.7B-Instruct-v1.0 | 80.28 | #1075 |
| 4 | Hermes-4-14B | 77.46 | #751 |
| 5 | Qwen 3 VL 8B Instruct | 76.06 | #420 |
| 6 | Qwen 3 VL 4B Instruct | 76.06 | #515 |
| 7 | Ministral-8B-Instruct-2410 | 76.06 | #1094 |
| 8 | Hermes-3-Llama-3.1-8B | 76.06 | #997 |
| 9 | Qwen 3 4B 2507 Instruct | 74.65 | #762 |
| 10 | AFM-4.5B | 74.65 | #1372 |
| 11 | Mistral-NeMo-Minitron-8B-Instruct | 73.24 | #907 |
| 12 | Llama 3.1 8B Instruct | 69.01 | #1035 |
| 13 | Apertus-8B-Instruct-2509 | 69.01 | #1061 |
| 14 | Granite 3.3 8B Instruct | 69.01 | #825 |
| 15 | GPT-OSS-20B | 67.61 | #518 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=latamboard-spanish-wnli · How It Works · Data refreshed daily, snapshot 2026-10-20.