Microsoft Introduces MAI-Thinking-1, a New Reasoning Model
Microsoft AI has launched MAI-Thinking-1, a medium-sized reasoning model with 35 billion active parameters and approximately 1 trillion total parameters (sparse Mixture of Experts model). Trained from scratch on enterprise-grade, commercially licensed data, it matches leading models like Claude Opus 4.6 on software engineering benchmarks and has advanced mathematical reasoning, achieving 97.0% on AIME 2025. It also scored higher than Claude Sonnet 4.6 in blind human evaluations.
MAI-Thinking-1 signifies Microsoft's direct entry into developing competitive in-house reasoning models, reducing reliance on third-party solutions. Its efficiency and strong performance for its size aim to make advanced AI assistance more accessible and integrated into daily enterprise workflows.
Learn one new AI thing every day.
Daily Deck sends you seven plain-English cards like this every morning. Free.
Start free