Microsoft launches three in-house AI models – signals independence from OpenAI

Microsoft today launched three new in-house AI models on its Foundry platform: MAI-Transcribe-1, MAI-Voice-1, and MAI-Image-2. The launch sends a clear message that the company wants to reduce its dependence on OpenAI, which has long been its primary AI partner.

MAI-Transcribe-1 is a speech recognition model competing directly with OpenAI's Whisper. MAI-Voice-1 is a text-to-speech model, while MAI-Image-2 is Microsoft's answer to image generation. All three are available via Azure Foundry from today, April 3, 2026.

The reasoning behind the launch is Microsoft's desire to control its own AI destiny. Although the company has invested heavily in OpenAI and integrated ChatGPT technology into Copilot, that dependency has created vulnerability. With its own models, Microsoft can offer AI services to enterprise customers without going through a third party.

Microsoft simultaneously announced its 2026 Release Wave 1 for Dynamics 365, Power Platform, and Microsoft 365 Copilot, with a strong emphasis on agentic AI. The company's ambition is clear: AI agents that act on behalf of users, not just answer questions.

For CIOs and IT leaders, this is significant news. Microsoft's platform is becoming stronger and more self-sufficient, offering better price stability and fewer dependencies in the supply chain.