Just ai-text to image there's adobe, dalle and a bunch of randoms. But you really need to play around and enginer the prompt well.