Microsoft Launches MAI-Image-1 for Image Generation

Microsoft has launched its first image generation model based on text prompts, MAI-Image-1, introduced in October 2025. Users can test it in Bing Image Creator or Copilot Audio Expression.
This marks Microsoft’s third proprietary model. In August 2025, the company released MAI-Voice-1 for natural speech creation and MAI-1-preview, its initial text model.
In Bing Image Creator, MAI-Image-1 is now available alongside DALL‑E 3 and GPT‑4o in the menu. In Copilot Audio Expression, it will be utilized in Story Mode for visualizing stories.
Currently, MAI-Image-1 is accessible in all countries where Bing Image Creator and Copilot Labs are available, including Ukraine, but it is not yet accessible in the European Union.
As previously stated by Microsoft, MAI-Image-1 "excels at generating photorealistic images, such as lighting effects (e.g., reflections, glare), landscapes, and more." The model has also ranked in the top 10 for text-based image generation on LMArena.