Microsoft has launched MAI-Thinking-1, its first in-house advanced reasoning AI model, developed from the ground up without relying on third-party models. This medium-sized model, featuring a sparse Mixture of Experts architecture with 35 billion active parameters, demonstrates competitive performance on software engineering benchmarks, matching leading models like Claude Opus 4.6 on SWE-Bench Pro. Microsoft also introduced six other AI models focused on image generation, transcription, voice, and coding, all trained on licensed data with internal infrastructure. AI
IMPACT Positions Microsoft as a direct competitor in advanced reasoning models, potentially influencing enterprise adoption and developer tool integration.
RANK_REASON Microsoft's announcement of a new, in-house developed frontier model with specific benchmarks.
- GitHub Copilot
- MAI-Code-1
- MAI-Image 2.5
- MAI-Thinking-1
- MAI-Transcribe-1.5
- MAI-Voice-2
- Microsoft
- OpenAI
- Visual Studio Code
- Build 2026
- Claude Opus 4.6
- SWE-Bench Pro
AI-generated summary · Google Gemini · from 4 sources. How we write summaries →