SmolLM-135M: benchmark results
Hugging Face's SmolLM-135M, the smallest of three SmolLM sizes (135M, 360M, 1.7B), trained on the Cosmo-Corpus dataset combining synthetic textbooks, Python code and educational web text. Provider: HuggingFace. Released 2024-07-16. Access: Open.
Aggregate Rating 1213 ± 41, rank #1577 of 1661 rated models, from 35 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - MuSR | 43.66 | Score | 75.5 |
| AI Energy Score (Text Generation) | 4 | Energy Score (1-5) | 47.5 |
| Open PL LLM - PSC (generative, 5-shot) | 72.48 | Binary F1 (%) | 21.9 |
| Open PL LLM - CBD (multiple choice, 5-shot) | 15.46 | Macro F1 (%) | 18.4 |
| Open Portuguese LLM - FaQuAD NLI | 43.97 | Macro F1 (%) | 17.1 |
| Open Portuguese LLM - ENEM | 21.13 | Accuracy (%) | 16.5 |
| Open Portuguese LLM - OAB Exams | 24.83 | Accuracy (%) | 14.6 |
| Open Portuguese LLM - ASSIN2 RTE | 44.37 | Macro F1 (%) | 13.9 |
| Open LLM Leaderboard - IFEval | 21.25 | Score | 13.7 |
| Open PL LLM - CBD (generative, 5-shot) | 15.08 | Macro F1 (%) | 11.6 |
| Open LLM Leaderboard - GPQA | 25.84 | Score | 10.5 |
| Open PL LLM - PPC (generative, 5-shot) | 33.6 | Accuracy (%) | 10.2 |
Interactive version: theaggregate.ai/model?slug=smollm-135m · How It Works · Data refreshed daily, snapshot 2026-10-23.