Thoughts on new model and maybe even changing again.
So i'm curious, not sure if Dev and mods is very active around the forums, usually just stick to text-to-image gen, but been floating around youtube, and apparently a short while ago a new AI model was released called "z-image" which seem to do stuff even better than Flux in most ways, do we know if the Dev have any plans to change it again? or sticking with the current Flux?
3 replies
i just wish it revert back to pre 8 august "nerf".
Is that when we changed from SD to Flux? (I personally prefer Flux, works better with my prompting style) or just when it got way slower?
Pre 8 august 2025, the image gen already use flux, me too like flux better than SD, but in my preference opinion, back then (pre 8/8/25) the quality are just way better, better accuracy, plenty artist style or library of characters are recognisable.
Now things seems to be more toward anime stuff, try "wednesday addams", "ramona flowers" or even "princess peach", how many image generated actually resembles the said characters.