LA Leaderboard - MGSM Spanish: leaderboard
MGSM grade-school maths word problems in Spanish, answered directly without chain of thought. From the LA Leaderboard, which evaluates models on Spanish, Catalan, Basque and Galician tasks with the lm-evaluation-harness; exact match after whitespace removal (%), higher is better. Each model result file is a separate run.
Metric: Exact Match (%). Source: huggingface.co. Saturation forecast: Not forecast. 69 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 3B | 18 | #1177 |
| 2 | Qwen3-8B-Base | 16.8 | #626 |
| 3 | SmolLM3-3B-Base | 15.6 | #1260 |
| 4 | OLMo-2-1124-7B-Instruct | 13.6 | #1343 |
| 5 | OLMo-2-1124-7B | 10 | #1381 |
| 6 | SmolLM3-3B | 9.2 | #1239 |
| 7 | gemma-3-4B-pt Base | 8 | #1110 |
| 8 | Gemma 3 4B (IT) | 6 | #997 |
| 9 | Yi-1.5-9B | 6 | #1215 |
| 10 | Qwen 2.5 7B | 5.2 | #895 |
| 11 | Gemma 2 9B | 4.8 | #924 |
| 12 | Qwen 2.5 1.5B | 4.4 | #1412 |
| 13 | SmolLM2-1.7B-Instruct | 4.4 | #1541 |
| 14 | Llama 3.2 3B | 4 | #1376 |
| 15 | EuroLLM-9B | 3.6 | #1238 |
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-mgsm-spanish · How It Works · Data refreshed daily, snapshot 2026-11-05.