Microsoft introduced its first proprietary image generator

Microsoft has officially unveiled MAI-Image-1, the first text-to-image generator developed entirely by its own team. This follows the launch of MAI-Voice-1 and MAI-1-preview models, which marked the beginning of a strategy to reduce reliance on the OpenAI partnership.

Microsoft says that MAI-Image-1 was developed with input from creative professionals to avoid "repetitive or overly formulaic results." The model is designed for photorealistic images, including natural scenes, lighting, and landscapes. It also runs faster on "larger, slower systems," making it suitable for integration into high-performance products.

MAI-Image-1 has already debuted in the top 10 of LMArena, a platform where users compare the results of different models and vote for the best ones. Microsoft plans to integrate the new generator into Copilot and Bing Image Creator in the near future.

The company emphasizes that it has paid special attention to safety and responsible use. Although independent tests are still ahead, Microsoft states its intention to ensure control of content and minimize the risks of abuse.