Microsoft Announces MAI-Image-1, First Self-made Image Making AI Model

JAKARTA - Microsoft introduced MAI-Image-1, an image-generating artificial intelligence (AI) model from its first text order, which was fully self-developed by Microsoft AI's internal team.

In its official announcement, Microsoft said that the MAI-Image-1 was designed for real value for creators, by avoiding repetitive or generically stylish image results.

The development of this model is also carried out through rigorous data selection and in-depth evaluation, based on input from creative industry professionals. So that it can provide flexibility and visual diversity for users.

MAI-Image-1 excels in producing photorealistic images, such as lighting (for example, reflections of light, reflections), landscapes, and more,' said the tech giant.

According to him, the combination of image creation speeds and quality allows users to display their ideas on the screen faster, compared to other models.

Microsoft also emphasizes its commitment to safe and responsible AI use. For this reason, MAI-Image-1 is currently being tested on the LMArena platform to collect feedback before it is widely launched.

We are committed to ensuring a safe and responsible result. This prompted us to start testing this model at LMArena so that we can gather insights and input, "said Microsoft.

The model will soon be available on Copilot and Bing Image Creator, two major services that have been integrated with Microsoft AI technology.