SmolLM3-3B-Base: benchmark results
Provider: HuggingFace. Released 2025-07-08. Access: Open.
Unified ELO 1319 ± 13, rank #1743 of 2096 rated models, from 330 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LA Leaderboard - MGSM Spanish | 15.6 | Exact Match (%) | 96.9 |
| LA Leaderboard - MGSM Catalan | 13.6 | Exact Match (%) | 95.4 |
| LA Leaderboard - MGSM Galician | 13.6 | Exact Match (%) | 93.8 |
| LA Leaderboard - OpenBookQA Spanish | 36.2 | Accuracy (%) | 91.2 |
| LA Leaderboard - TELEIA | 57.14 | Accuracy (%) | 91.2 |
| EuroEval German Summarization - Mlsum DE | 37.45 | Score (%) | 90.5 |
| LA Leaderboard - Summarization Galician | 6.88 | BLEU | 89.6 |
| LA Leaderboard - RagQuAS | 80.6 | Semantic Answer Similarity (x100) | 88.2 |
| LA Leaderboard - xStoryCloze Spanish | 70.09 | Accuracy (%) | 85.3 |
| LA Leaderboard - OpenBookQA Galician | 28.4 | Accuracy (%) | 83.8 |
| EuroEval German NLU - Germanquad | 61.5 | Reading comprehension Score (%) | 82.8 |
| LA Leaderboard - caBREU Catalan | 15.54 | BLEU | 82.4 |
Interactive version: theaggregate.ai/model?slug=smollm3-3b-base · How It Works · Data refreshed daily, snapshot 2026-10-08.