Q: Several suggestions and question
First, Thanks for what you built so far.
The platform has strong potential and there are hidden gems but there are some things to polish/improve. Remove models with bad prompt adherence or move them to an “Old Models/Bad Prompt Adherence” category. Honestly, any model with poor prompt adherence should be removed. Keep just the models that have good prompt adherence, some good models are —Flux, Flux Anime, Flux Realism, Analog Madness, cute wave, shuttle aesthetic, Face Bomb, Babes, and all Super Machine series. I know you are trying to make the platform look great with lots of models but removing or isolating models with bad prompt adherence will result in a better user experience.
questions:
Where is the instruct edit, flux img to img option, style assistance, lora's option , it was in the old layout but can't find in the new layout?
Rexxar
Jun 9, 2025A: Thanks for your feedback. The instruct edit is more built in now on the new version, click any image and it auto puts you into the edit view. With the edit, it works better than the flux img 2 img did so that has taken a backseat.
Style assistance is also more integrated now with the prompt enhancer toggle in the advanced settings. LORA UX is coming shortly and the team UX should be live in the next 2 days.
In terms of models, agree we have a lot and some are older than others. We'll look into the best way to categorize them for simple use. There is a whole video series on the new UX underway where we will go into some of the models in more depth also.
Thanks!