hi Karen Majercsik, yes unfortunately this is a limitation/challenge with almost all AI image models. I would try different prompts, switching models, or explicitly adding "no text" to the prompts