Ideogram is working on a new text-to-image model 3.0 (beta). Comes with even better text handling

→ View original post on X — @testingcatalog