Mistral Medium 3.5 (High) — benchmark results

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

Unified ELO 1623 ± 22, rank #425 of 1752 rated models, from 13 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
PACT (Lechmazur)1538CMS Points60
Finance Agent v1.146.11Score (self-reported)29.1
Vals AI ProofBench10Accuracy (%)24.7
MedCode33.75Score (self-reported)20.8
Buyout Game (Lechmazur)1332.2Bradley-Terry Rating20
NYT Connections Extended27.8Score (%)17.4
Finance Agent v232.1Score (self-reported)16.7
Multi-turn Debate (Lechmazur)1398.1Bradley-Terry Rating14.7
MedScribe67.73Score (self-reported)9.6
Vals Index37.81Score (self-reported)8.3
Vals Multimodal Index34.77Score (self-reported)5.3
Sycophancy (Lechmazur)22.1Sycophancy rate % (lower is better)3.2

Interactive version: aibenchmarks.dev/model?slug=mistral-medium-3-5-high · How the rankings work · Data refreshed daily, snapshot 2026-07-16.