Microsoft unveils its first reasoning model and six other AI tools at Build

Microsoft unveils its first reasoning model and six other AI tools at Build">

Microsoft kicked off its annual Build developer conference on Tuesday with a keynote, during which the company announced seven new AI models, including its first reasoning model. The new models are part of Microsoft's push to develop more advanced AI capabilities, with a focus on enterprise-grade applications. Microsoft AI CEO Mustafa Suleyman introduced the new models, emphasizing the company's "humanist superintelligence" framing.

The first reasoning model, MAI-Thinking-1, was trained on "enterprise-grade, clean and commercially licensed data," a key consideration given mounting concerns about copyright and AI use. The 35-billion-parameter model reportedly beat Anthropic's Sonnet 4.61 in a blind test and aligns with Anthropic Opus 4.6 in its SWE Bench Pro benchmark score for coding. In addition to MAI-Thinking-1, Microsoft announced several other new AI models, including MAI-Code-1, an "ultra-efficient" coding model tuned for GitHub; MAI-Image-2.5, a text-to-image and image-to-image model; MAI-Transcribe-1.5, a transcription model that combines state-of-the-art accuracy across 43 languages; and MAI-Voice-2, a voice model with improved cost efficiency.

All the new MAI models will be available on Fireworks AI, Baseten, and Open Router. Suleyman emphasized the security frameworks of the new models, noting that "everything is watermarked from scratch." He also highlighted cost efficiency improvements across each model, with some showing as high as 10x improvements compared to similar competitor models. The company has already seen significant interest in its AI models, with MAI-Image-2.5 hitting the third spot on the LM Arena Leaderboard.

The announcements also included a collaboration with the Mayo Clinic to develop a new frontier model for healthcare. The project joins the growing number of health-specific AI applications from companies including OpenAI and Google. Microsoft already offers Copilot Health, but data privacy, security, and hallucinations are concerns when it comes to medical AI.

The rapid pace of AI model releases highlights the intense competition in the field, with companies racing to develop and deploy more advanced capabilities. As Suleyman noted, the new models are designed to be more efficient and effective, with a focus on enterprise-grade applications.