Mistral Small 4 — benchmark results

March 2026 Mistral Small 4 model row. Provider: Mistral. Released 2026-03-16. Access: Open.

Unified ELO 1559 ± 22, rank #597 of 1776 rated models, from 46 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
SpeechMap Compliance90.9% Requests Completed91.4
UGI - Writing40.33Writing Score74.5
UGI - Natural Intelligence27.77NatInt Score69.4
YapBench320.2YapIndex (lower is better)60.2
Wolfram LLM Benchmarking Project43.1Correct Functionality (%)54.5
AraGen33.853C3H Score (%)47.8
ZeroEval GPQA Diamond71.2GPQA Diamond Score46.9
UGI Leaderboard33.58UGI Score46.2
AI for Education Pedagogy - Maths73.02Accuracy (%)43
IDP Leaderboard69.6Avg Score (%)40
RewardBench 2 Focus86.36Accuracy (%)39.2
MATH-MC Level 496.25Accuracy (%)36

Interactive version: theaggregate.ai/model?slug=mistral-small-4 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.