Mistral Large 2512: benchmark results

Provider: Mistral. Released 2025-12-02. Access: Open.

Unified ELO 1719 ± 29, rank #291 of 2656 rated models, from 14 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
TaxEval v273.06Score (%)64.9
Vals AI CaseLaw v261.41Accuracy (%)61.9
CorpFin v261.03Score (%)56.8
LegalBench79.14Score (%)37.5
MMLU Pro79.82Score (%)35.5
AIME42.92Score (%)31.6
GPQA Diamond68.43Score (%)30.4
MedQA82.23Score (%)29.8
MMMU Pro66.19Score (%)22.4
IOI4Score (self-reported)20
MortgageTax52.11Score (%)16.9
LLM-SoccerArena144Prediction Points16.7

Interactive version: theaggregate.ai/model?slug=mistral-large-2512 · How It Works · Data refreshed daily, snapshot 2026-09-19.