Mistral Small 3.1 24B Instruct: benchmark results

Provider: Mistral. Released 2025-03-17. Access: Open.

Unified ELO 1526 ± 1, rank #568 of 1392 rated models, from 26 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
MERA - ruDetox35.55Joint Score (%)92.8
MERA - MultiQ56.87F1 (%)80.9
LLM Stats (TriviaQA)80.5Score (%)79.4
MERA - PARus91.2Accuracy (%)78.9
MERA - ruTiE86.07Accuracy (%)78.7
MERA - USE35.39Grade, normalized (%)73.2
MERA - MaMuRAMu81.33Accuracy (%)72.2
MERA - CheGeKa32.4F1 (%)71.3
MERA - SimpleAr99.3EM (%)70.9
MERA - ruOpenBookQA90.25Accuracy (%)70.8
MERA - ruMMLU71.4Accuracy (%)68.9
MERA - RCB56.39Accuracy (%)68.4

Interactive version: theaggregate.ai/model?slug=mistral-small-3-1-24b-instruct · How It Works · Data refreshed daily, snapshot 2026-09-05.