I’ve been testing some image-generation AIs and I have to admit I’m pretty disappointed. The AI hallucinates a lot. It doesn’t read the whole prompt, so I end up having to send a bunch of prompts for a single image that, to me, isn’t even worth using, because it’s so obvious it was made by AI. There are tons of flaws, especially with GPT Images and Flux.
The only one I’ve tested that I know is really worth it is Stable Diffusion. I even saw this thread here on the forum:
But it looks like it doesn’t work anymore 