MobileLLM-R1-360M: benchmark results
Meta MobileLLM-R1 360M is the final reasoning SFT checkpoint for mathematics, programming and scientific problems, separate from its Base release. The publisher specifies 15 layers, 16 attention heads, 4 key/value heads and shared input/output embeddings. Provider: Meta. Released 2025-09-12. Access: Open.
Aggregate Rating 1176 ± 38, rank #9735 of 10916 rated models, from 143 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Faroese NLU - ScaLA FO | 0.78 | Matthews correlation coefficient (%) | 30.9 |
| EuroEval Lithuanian Common Sense Reasoning | 0.51 | Common Sense Reasoning Average Score (%) | 26.5 |
| EuroEval Estonian NLU - Grammar ET | 0.58 | Matthews correlation coefficient (%) | 22.5 |
| EuroEval Icelandic Knowledge | 1.38 | Knowledge Average Score (%) | 21.3 |
| EuroEval Estonian Knowledge | 0.71 | Knowledge Average Score (%) | 19.5 |
| EuroEval Icelandic Common Sense Reasoning | 0.48 | Common Sense Reasoning Average Score (%) | 18.9 |
| EuroEval Norwegian Knowledge - Idioms NO | 0 | MCC (x100) | 18.6 |
| EuroEval Norwegian Knowledge - NRK Quiz QA | 2.9 | MCC (x100) | 15.6 |
| EuroEval Lithuanian Knowledge | 1.2 | Knowledge Average Score (%) | 15.2 |
| EuroEval Norwegian Knowledge | 1.45 | Knowledge Average Score (%) | 14.5 |
| EuroEval Icelandic NLU - Hotter and Colder Sentiment | 3.43 | Matthews correlation coefficient (%) | 14.4 |
| EuroEval Czech Common Sense Reasoning | 0.27 | Common Sense Reasoning Average Score (%) | 14 |
Interactive version: theaggregate.ai/model?slug=mobilellm-r1-360m · How It Works · Data refreshed daily, snapshot 2026-11-08.