Microsoft
ModelsModels from this lab.16
RankedModels with an index.3
Top modelModel with the highest index.53.5MAI-Thinking-1
LatestRelease of the newest model.2026-10Microsoft-Decision-1
Best by category
Out of 100AgenticMulti-step tasks with tools.—
CodingCode writing and repair.MAI-Thinking-155.2
ReasoningLogic problems and puzzles.Phi-4-reasoning22.1
MultimodalTasks with images and text.—
KnowledgeFacts and expert knowledge.MAI-Thinking-153.0
MultilingualTasks in many languages.Phi-424.0
InstructionTasks with strict prompt rules.Phi-4-reasoning21.1
MathMath problems.MAI-Thinking-152.6