Microsoft AI has released MAI Image 2.7 in preview. Unusually, the announcement calls it “our latest image editing model”: “We’re starting with editing, with text-to-image generation coming soon.”
What it is for
Microsoft pitches it at the hard edits: restyling or relighting a whole scene, changing text and symbols “while keeping them integrated into the image”, restoring old assets, and moving a product into a new setting without breaking the lighting.
The rankings
Microsoft says MAI Image 2.7 launched at No. 3 on Arena’s image-edit leaderboard, which makes Microsoft AI the No. 2 lab there, and No. 2 on Artificial Analysis’s image-editing leaderboard, above OpenAI’s GPT Image 2.5 Flare. It says the model improves on 2.6 “across all 10 use cases and all 7 editing actions evaluated by Artificial Analysis”, with its biggest lead in scene and style, “24 Elo ahead of the next-best model”.
Treat ranks as a snapshot: these boards move week to week, and these are Microsoft’s launch-day figures.
Where to try it
The post points to Arena, Artificial Analysis and Microsoft’s own MAI Playground. It does not mention pricing, an API or Azure availability yet. For how the main image models compare in practice, see our Nano Banana vs GPT Image guide.
The Campfire
CommentsThanks. Your comment is waiting for the desk. Once it is approved it appears here within about a minute.
Add a comment