I was waiting for any models better than sdxl to release. Something trained on booru data. There are videos now, but besides that quality has stagnated more than I expected.
that newest chatgpt image generator is so much better than anything stable diffusion can do right now. looking forward to what people make. If you make something or have a good prompt please show me! the wait times are pretty bad right after releasing so its hard to experiment
@mindovertease I don't know if I can compare it to illustrious or other local models. It can't do NSFW or a lot of the stuff you can do locally. Different use-cases. But the quality is really good, and almost every generation is really solid. But takes like 1-2 minutes each.