MAI-Image-1: Microsoft’s first in-house AI image generator debuts

Microsoft has introduced MAI-Image-1, its first internally developed AI system for generating images from text, marking a step in the company’s strategy to build more of its own models alongside its partnerships. 

The model is available for public testing on the LMArena benchmarking platform and is slated for integration into Copilot and Bing Image Creator, the company said. 

Designed with feedback from creative professionals, MAI-Image-1 aims to avoid generic outputs and emphasises photorealistic rendering, with strengths in elements such as lighting, landscapes and scene composition. 

Microsoft also claims the model runs faster than some of its larger rivals. Early community scoring placed it among the higher-ranked image models on LMArena. 

The debut complements Microsoft’s broader push to develop purpose-built AI, alongside work with partners like OpenAI and Anthropic, and follows the company’s recent disclosure of other in-house models, including MAI-1-preview and MAI-Voice-1.