mistral-small-latest: benchmark results
Rolling alias for Mistral's current Small - as of mid-2026 Mistral Small 4, a 119B MoE (6.5B active) hybrid unifying chat, reasoning and coding (256K context). Provider: Mistral. Released 2026-03-16.
Aggregate Rating 1545 ± 43, rank #1766 of 10916 rated models, from 14 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Arcadia MMMU Open Ended | 35.85 | Accuracy (%) | 90 |
| CyberSecEval2 Interpreter Abuse | 18.55 | Malicious Compliance Rate (%) | 77.8 |
| CyberSecEval2 Prompt Injection | 28.59 | Injection Success Rate (%) | 55.6 |
| Enkrypt AI - Safety Risk | 28.47 | Risk Score | 45.1 |
| Enkrypt AI - Jailbreak Risk | 10.53 | Risk Score | 43.5 |
| Enkrypt AI - Bias Risk | 85.79 | Risk Score | 37 |
| Arcadia MMMU Multiple Choice | 54.55 | Accuracy (%) | 36.4 |
| OpenAI HumanEval | 82.32 | Accuracy (%) | 31.8 |
| Enkrypt AI - Toxicity Risk | 5.45 | Risk Score | 30.1 |
| Enkrypt AI - Insecure Code Risk | 52.44 | Risk Score | 20.3 |
| CyberSecEval2 Vulnerability Exploit | 35.91 | Accuracy (%) | 20 |
| Enkrypt AI - Risk Score | 43.22 | Risk Score | 15.9 |
Interactive version: theaggregate.ai/model?slug=mistral-small-latest · How It Works · Data refreshed daily, snapshot 2026-11-08.