Noesion Radarnoesion.ai
Release2026-08-12

Microsoft released MAI-Thinking-1, its own reasoning model, in public preview on Microsoft Foundry. Microsoft says it was trained without copying outputs from other labs' models, matches Claude Opus 4.6 on a software engineering benchmark, and was preferred to Claude Sonnet 4.6 by paid human raters. It is a preview, and the comparison figures come from Microsoft's own testing.

What changed

Microsoft's in-house MAI family had no dedicated step-by-step reasoning model of this size.

What it unlocks

Testing a Microsoft-built reasoning model for coding and maths work through Microsoft Foundry, with a context window large enough for very long documents.

  • 35B active, ~1T total parameters
  • 97.0% on AIME 2025
  • 256k token context window
  • 1,276 tasks in human evaluation

What you need to act on it

  • Microsoft Foundry access
  • public preview

Sources

Send this to someone who needs it

Shares the story and its sources. Nothing about you.

What does this mean for your job?

This is the story as everyone gets it. Once a week we send you the version written for your role — what changed, why it matters for the work you actually do, and one thing to try. Free while we tune it.