Devstral Small (2505): benchmark results
Provider: Mistral. Released 2025-05-01. Access: Open.
Aggregate Rating 1530 ± 26, rank #1954 of 10916 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Twinkle Eval - TW-Legal Benchmark v1 | 44.57 | Accuracy (%) | 64.3 |
| Twinkle Eval - TMMLU+ | 60.09 | Accuracy (%) | 53.6 |
| Twinkle Eval - MMLU | 75.3 | Accuracy (%) | 50 |
| Wolfram LLM Benchmarking Project | 35.2 | Correct Functionality (%) | 35.7 |
| AI Chess Leaderboard (Reasoning) | 599 | Elo | 33 |
| Multi-LCB - Pass@5 T0.6 - Go | 28.5 | Pass@5 (%) | 25 |
| Multi-LCB - Pass@5 T0.6 - Rust | 29.1 | Pass@5 (%) | 20.8 |
| Multi-LCB - Pass@5 T0.6 - Scala | 25.2 | Pass@5 (%) | 20.8 |
| Multi-LCB - Pass@5 T0.6 - TypeScript | 33.3 | Pass@5 (%) | 20.8 |
| Kagi LLM Benchmark | 37.7 | Accuracy (%) | 18.7 |
| Multi-LCB - Pass@5 T0.6 - JavaScript | 30.8 | Pass@5 (%) | 16.7 |
| Multi-LCB - Pass@5 T0.6 - PHP | 29.5 | Pass@5 (%) | 16.7 |
Interactive version: theaggregate.ai/model?slug=devstral-small-2505 · How It Works · Data refreshed daily, snapshot 2026-11-08.