Microsoft unveils new AI models to lessen reliance on OpenAI and lower costs

What happened

At Microsoft Build, Microsoft unveiled MAI-Code-1-Flash, its first in-house MAI-series model focused on coding workloads. It is positioned as a cheaper alternative for developers on Azure.

Context and impact

The move continues Microsoft's diversification — after a year of investing in Anthropic, Mistral and its internal MAI team, it is reducing dependency on OpenAI exclusivity.

Details

• launch: Microsoft Build 2026, June 2
• positioning: "flash" tier, optimized for latency and cost
• target: developers in Azure Foundry and the GitHub Copilot stack