MAI-Image-2.5 launches at No. 3 on Arena text-to-image leaderboard

Today, we're announcing MAI-Image-2.5 - ranked third on the Arena text-to-image leaderboard.

It's our strongest image model yet, and the next step in the MAI-Image series.

MAI-Image-2.5 performs well across a wide range of styles, it follows instructions closely, renders text more reliably than ever, and produces detailed, coherent images, just as you intend.

This is a step change in quality over our previous model, MAI-Image-2, delivering major improvements in text rendering, stylized illustration and commercial imagery.

The model also shows strong visual reasoning across objects, scene structure, lighting, scale, and spatial relationships - helping turn simple directions into polished images.