Mistral Small 4: benchmark results
Mistral AI's Small 4, a 119B sparse MoE (6.5B active) model (March 2026). Provider: Mistral. Released 2026-03-16. Access: Open.
Unified ELO 1551 ± 1, rank #428 of 1392 rated models, from 100 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SpeechMap Compliance | 90.9 | % Requests Completed | 92.2 |
| StereoTales | 117 | Emissions (self-reported) | 90.9 |
| Guesswork 2026-07 | 0.93 | MAE (z-score units) | 75 |
| UGI - Writing | 40.33 | Writing Score | 73.8 |
| UGI - Natural Intelligence | 27.77 | NatInt Score | 68.7 |
| BLXBench | 75.2 | Score (self-reported) | 66.7 |
| Wolfram LLM Benchmarking Project | 43.1 | Correct Functionality (%) | 52.6 |
| Conceptual Reasoning Index - Decision Theory (DTBench) | 51.55 | Chance-Corrected Score (0-100) | 51.3 |
| SEA-HELM (Tamil) | 55.09 | Mean Score | 51.2 |
| SEA-HELM (Filipino) | 56.18 | Mean Score | 48.8 |
| AraGen | 33.85 | 3C3H Score (%) | 47.8 |
| Epoch AI - Dtbench | 70.93 | Score | 46.9 |
Interactive version: theaggregate.ai/model?slug=mistral-small-4 · How It Works · Data refreshed daily, snapshot 2026-09-05.