Magistral Medium 1.2 vs Grok 4.20

Best cell per row highlighted. Null means undisclosed — never counted as zero.

Mistral AI
Magistral Medium 1.2GA
7.2
AI Panel
xAI
Grok 4.20GA
6.9
AI Panel

Identity & lifecycle

ProviderMistral AIxAI
Family / tierMagistral · ReasoningGrok 4 · Reasoning
StatusGAGA
Released2025-09-172026-03-09
Knowledge cutoff2025-062024-11

Architecture & context

Context window131K1M
Max output tokens41K
Input modalitiestext, imagetext, image
Reasoning modealwaysoptional
Open weightsNoNo

Pricing (per Mtok)

Input$2$1.25
Output$5$2.5
Cached input$0.2
Batch input$1
Free tierYesYes

Speed

Output speed (tok/s)38.9171.4
Time to first token (s)1.713.24
Latency tierslowmedium

Trust & deployment

Trains on inputsNoYes
LicenseProprietaryProprietary
CloudsAzure AI Foundry
ComplianceSOC2, ISO27001, GDPR

AI Panel scoring

Unified score7.26.9
Decision Maker77
Domain Strategist6.56.5
Finance Lead76.5
Domain Practitioner7.57.5
Power User7.56.5
Skeptic6.56
Value score77

Benchmarks

GPQA Diamond76.2678.5
AIME 202583.48
MATH-50087.3
MMMU70
IFEval81
TAU-bench93
LMArena Elo1491
Artificial Analysis Index2749

Scores link to their sources. Missing cells mean the vendor hasn't disclosed a result — honesty over padding.