priorsBuilder

Mistral · released 2025-07-10

Devstral Medium

Measured on 15 benchmarks.

General frontier

Not ranked in this mix — too little measured data for its benchmarks.

Your mix’s benchmarks

2 of 8 measured

reasoning

Humanity's Last Exam

587 results
  1. #548none of 5873.8%

Leader: Opus 5.5 · 61.4%

knowledge

AA-Omniscience

526 results
  1. #278none of 526-31.6

Leader: Opus 5.5 · 46.4

All 15 benchmarks

show 13 more

agentic

τ²-Bench Telecom

393 results
  1. #323none of 39319.9%

Leader: GLM 5.2 · 99.1%

coding

LiveCodeBench

282 results
  1. #170none of 28233.7%

Leader: Gemini 3 Pro · 91.7%

coding

Terminal-Bench Hard-AA

387 results
  1. #230none of 3879.1%

Leader: GPT-5.6 Sol · 65.9%

composite

AA Intelligence Index v4.3.2

610 results
  1. #400none of 6109.0

Leader: Opus 5.5 · 57.6

instruction

IFBench

399 results
  1. #343none of 39929.9%

Leader: Grok 4.3 · 83.3%

knowledge

AA-Omniscience Accuracy

526 results
  1. #313none of 52618.9%

Leader: Fable 5.1 · 67.2%

knowledge

AA-Omniscience Non-hallucination

526 results
  1. #168none of 52637.8%

Leader: MiniCPM5-1B (Non-reasoning) · 99.1%

knowledge

GPQA Diamond

539 results
  1. #417none of 53949.2%

Leader: GPT-6 Astra · 96.3%

knowledge

MMLU-Pro

285 results
  1. #176none of 28570.8%

Leader: Gemini 3 Pro · 89.8%

long-context

AA-LCR v1.1

523 results
  1. #369none of 52332.0%

Leader: Kimi K3 · 88.7%

math

AIME 2025

228 results
  1. #210none of 2284.7%

Leader: GPT-5.2 · 99.0%

math

MATH-500

162 results
  1. #113none of 16270.7%

Leader: GPT-5 · 99.4%

reasoning

CritPt

529 results
  1. #401none of 5290.0%

Leader: Muse Wonder Full (Max) · 32.9%