QwQ-32B vs Magistral Medium 1.2

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

Alibaba Cloud
QwQ-32BGA
6.8
AI Panel
Mistral AI
Magistral Medium 1.2GA
7.2
AI Panel

Identity & lifecycle

ProviderAlibaba CloudMistral AI
Family / tierQwQ · ReasoningMagistral · Reasoning
StatusGAGA
Released2025-03-042025-09-17
Knowledge cutoff2024-092025-06

Architecture & context

Context window131K131K
Max output tokens33K41K
Input modalitiestexttext, image
Reasoning modealwaysalways
Open weightsYesNo

Pricing (per Mtok)

Input$0.12$2
Output$0.18$5
Cached input
Batch input$1
Free tierYesYes

Speed

Output speed (tok/s)38.9
Time to first token (s)1.7
Latency tierslowslow

Trust & deployment

Trains on inputsNoNo
LicenseApache-2.0Proprietary
CloudsGCPAzure AI Foundry
ComplianceSOC2, ISO27001, GDPR

AI Panel scoring

Unified score6.87.2
Decision Maker77
Domain Strategist76.5
Finance Lead77
Domain Practitioner7.57.5
Power User67.5
Skeptic6.56.5
Value score87

Benchmarks

GPQA Diamond65.276.26
AIME 202583.48
MATH-50090.6
LiveCodeBench63.4
MMMU70
IFEval83.9
Artificial Analysis Index27

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