Microsoft AI has announced MAI-Image-1, its first internally developed image generation model. The model debuts among the top 10 text-to-image models on the LMArena leaderboard. MAI-Image-1 focuses on delivering high-quality, photorealistic imagery, and is designed to be flexible and efficient for creators. Microsoft says the model is trained with care to avoid repetitive or generic outputs, and they plan to deliver genuine value for creators, and put a lot of care into avoiding repetitive or generically-stylized outputs. Integration and Availability MAI-Image-1 is now available in select Microsoft products. As of November 4, 2025, the model has begun launching in the following experiences:

Bing Image Creator: MAI-Image-1 is available as an option alongside other models, accessible at bing.com/create, within the Bing mobile app, and directly from the Bing search bar.
Copilot Audio Expressions: When selecting Story Mode, Audio Expressions will utilize MAI-Image-1 to create a unique visual representation of the user's narrative.

Currently, MAI-Image-1 is available in all regions with access to Bing Image Creator and Copilot Labs, excluding the EU.