Microsoft AI's First In-House Image Generator, MAI-Image-1, Now Available

Microsoft has launched its first in-house AI image generator, MAI-Image-1, in two of its products: Bing Image Creator and Copilot Audio Expressions. The company initially announced the model in October. According to a post by Mustafa Suleyman, CEO of Microsoft AI, on X, the text-to-image model will be “coming soon” to the European Union.

Suleyman added that the model excels at generating images of food and natural scenes, as well as those featuring artistic lighting and photorealistic details.

Earlier, Microsoft shared more insights on its blog: “MAI-Image-1 is particularly strong at producing realistic imagery—such as lighting effects (e.g., reflected light and highlights) and landscapes. This holds especially true when compared to many larger, slower models. Its combination of speed and quality allows users to visualize their ideas more quickly, iterate rapidly, and then move their work into other tools for further refinement.”

Additionally, MAI-Image-1 will power AI-generated artwork in the “Story Mode” of Copilot Audio Expressions—a text-to-speech platform within Copilot—creating visual companions for AI-narrated audio stories.

In August, Microsoft unveiled its first internal AI models: the voice model MAI-Voice-1 and the text-based MAI-1-preview. At the time, the company indicated it planned to integrate MAI-1-preview into its Copilot AI assistant in certain unspecified scenarios, signaling a potential shift away from reliance on OpenAI’s models. As of now, however, Microsoft states that its Copilot chatbot is transitioning to OpenAI’s newest model, GPT-5, while also offering users access to Anthropic’s Claude AI as an alternative.

MAI-Image-1 is currently listed as one of three available AI models on the Bing Image Creator website and app. The other two—DALL·E 3 and GPT-4o—are developed by OpenAI.