CLUSTER · TIER 3
Microsoft launches MAI-Thinking-1 reasoning model and MAI-Code-1-Flash for Copilot.
Microsoft announced two new LLMs: MAI-Thinking-1, a 35B reasoning model available to select early partners, and MAI-Code-1-Flash, a 5B model purpose-built for GitHub Copilot and VS Code. The company claims MAI-Thinking-1 is preferred to Sonnet 4.6. The low parameter counts stand out given current GPU scarcity.
Sources
1
X mentions
—
First seen
3Dago
Velocity
+2%/6h