Microsoft Launches Free AI Image Generator, MAI-Image-1, Integrated with Bing and Copilot
Havana, Cuba - November 26, 2025 – Microsoft has released MAI-Image-1, a new artificial intelligence model capable of generating photorealistic images, available to all users with a Microsoft account. The launch, reported by Cubadebate, integrates the tool directly into Bing and Copilot Audio expressions.
MAI-Image-1 distinguishes itself through its ability to create images with detailed photorealistic qualities, natural scenes, artistic lighting, and realistic food representations. Users can generate images by providing text descriptions or uploading existing images for AI-powered change. Developed internally by Microsoft, the model was trained on carefully selected data and refined with feedback from creative professionals.
According to Microsoft, the AI excels at handling complex visual details, offering excellent lighting, textures, and overall detail. While currently not at the level of platforms like Midjourney, early tests indicate solid performance, particularly with detailed prompts. Contextual elements like lighting and photographic technique are recommended for optimal results.
MAI-Image-1 supports image generation in 1:1, 3:2, and 2:3 aspect ratios, mirroring those used by GPT-4o.Downloadable images are limited to a maximum resolution of 1248 by 832 pixels. The system generates new images based on modified prompts or uploaded photos, and does not allow editing of previously created images.
Beyond Bing, MAI-Image-1 is also available within Copilot Audio Expressions, an experimental feature that generates images based on audio content analysis – such as, creating visuals to accompany themes or emotions within a Story mode episode.
