Compare AI models

Specifications
Parameters
671B—
Context window
128k128k
API pricing
Cheapest input
$0.25DeepInfra—
Cheapest output
$0.95DeepInfra—
$97.50DeepInfra · fp4—
Benchmarks
LiveCodeBench
74.8%75%
AIME 2024
93.1%91.82%
AIME 2025
88.4%83.48%
GPQA Diamond
80.1%76.26%
Overview
CompanyDeepSeekMistral
Release dateAug 21 2025Sep 17 2025
AccessOpen WeightClosed
Model detailsView modelView model

Frequently asked questions

DeepSeek V3.1 leads Magistral Medium 1.2 on 3 of the 4 benchmarks they both report (LiveCodeBench, AIME 2024, AIME 2025, GPQA Diamond). DeepSeek V3.1 shipped 27 days before Magistral Medium 1.2, so benchmark comparisons should account for the intervening progress.

Context windows are 128k (DeepSeek V3.1) vs 128k (Magistral Medium 1.2). DeepSeek V3.1 is open weight, while Magistral Medium 1.2 is closed.

On LiveCodeBench, Magistral Medium 1.2 leads at 75% vs DeepSeek V3.1 at 74.8%. On AIME 2024, DeepSeek V3.1 leads at 93.1% vs Magistral Medium 1.2 at 91.82%. On AIME 2025, DeepSeek V3.1 leads at 88.4% vs Magistral Medium 1.2 at 83.48%. On GPQA Diamond, DeepSeek V3.1 leads at 80.1% vs Magistral Medium 1.2 at 76.26%.