Feature Requests

allow a simple brush tool to erase portions of the image and run through the masked area into the algorithm again with the same prompt or ability to modify the prompt/style/any settings to regenerate the missing area

Your log-in still sucks! I've been trying for over half-an-hour tonight to log in and it just hangs on the "verifying" crap. Why am I paying for this????

Essential v1 has the feature, whereas Ev2 just recreates the sketch / drawing. Strangely, if an anime model generated by Ev2 is used as a base to generate Photo/Art there is no issue. But Ev1 is better in regard of recognizing and transferring / transforming content, especially people, to other models. PS: image to image

Currently only Image to Image can be used with Flux.1 [dev] in Essential when using image references. When will it be expanded to include options like Pose, Hard Edges, Soft Edges, Style etc...???

Add capacity to Train Lora Textual Inversion embedding. SO users can see the power of SD!

Runway has video to video it'd be great if it was on here

I deleted several files by accidentally clicking on the delete button, and I would like to restore them or undo the delete action. I think a section will be useful here, where deleted files will be stored for a certain period of time, so that accidentally deleted files can be restored

Is there any way to get better feedback on the model trainer for images it says are NSFW? I have an adult female fully clothed, sleeping, and yet it's marked as NSFW. As far as I know, there's no feedback other than it being flagged.