Microsoft Build 2026: Proprietary MAI-Code and MAI-Thinking Models Launched to Reduce OpenAI Reliance
At its Build 2026 developer conference in San Francisco, Microsoft announced a major strategic pivot by launching its own proprietary AI models. This move represents a concerted effort to compete directly with OpenAI, Anthropic, and Google, while reducing the massive costs associated with routing developer traffic to third-party models.
Microsoft introduced two primary models:
- MAI-Code-1-Flash: Microsoft's first model in the AI coding space. It is designed to be "inference ultra-efficient" and is integrated directly into GitHub Copilot and Visual Studio Code.
- MAI-Thinking-1: A medium-sized reasoning model available in private preview on Microsoft Foundry, designed to deliver high-performance reasoning at a low-token cost.
According to Mustafa Suleyman, CEO of Microsoft AI, refining these models for enterprise partners like McKinsey allowed Microsoft to outperform OpenAI's GPT-5.5 with 10 times better cost efficiency. This highlights Microsoft's transition from being primarily a cloud hosting partner and venture investor (having invested $13 billion in OpenAI and $5 billion in Anthropic) to a direct competitor at the model frontier.