Microsoft Introduces MAI-Image-2
Credit: microsoft.ai
Last October, Microsoft introduced its AI image-generation model, MAI-Image-1, which ranked ninth on the Arena.ai platform. Now, the software giant has announced the launch of its second-generation algorithm, MAI-Image-2, capable of creating images with more natural lighting, accurate skin tones, and realistic details.
MAI Image 2 features increased generation speed and improved quality compared to its previous version. It supports a wide range of styles, handles details, and correctly interprets complex text descriptions. The system implements content filtering and protection against the generation of inappropriate images.
The new version is integrated into Microsoft cloud services and is available to developers via an API. It supports scalability for enterprise workloads and integration with other company products. MAI Image 2 features optimizations to speed up output and reduce computational costs.
The model is trained on a large volume of heterogeneous data, enabling it to create high-resolution images that are context-sensitive and tailored to the specified parameters. The system includes tools for customizing the style, format, and resolution of the resulting image.
Copilot and Bing Image Creator users will be able to use the MAI-Image-2 model soon. The algorithm is currently available through the MAI Playground platform, and developers can access the model through the corresponding API on Microsoft Foundry.
