INPAINT

This is a multi use AI Tool - Upload two exact copies of the same Image - choose whether you want to Mask and Inpaint a specific area - or just Img2Img the whole thing, if you feel really advanced - the Img2Img switch just inverts the mask - so you could mask the face in that mode - and it would change everything else, pick how many Lora's you want to use - default is just the first 1, but you can turn on all 3 of them. Upscales by default to 2048 Pixel FB Server Ready, but adjust to taste. Denoise will set how much to allow the image (or area) to be changed - .5 will give subtle changes, .65 to .8 will give more dramatic changes. Changing Guidance will also effect the prompt adherence and realism of the output. I suggest suggest between 2 - 5 works well.

Since Flux.1 models are not deliberately intended for inpainting, this tool is experimental. Sometimes I've gotten better results than SDXL inpainting though, so here it is.You just might have to play a bit with guidance and denoise, and sometimes also switching model.

UPDATE Fixed an error in the workflow that occurred when changing the resolution of the image (which probably means 90% of the time).A tool I've created for my personal inpainting needs, but why not sharing it.It's originally intended for spot healing and correction techniques, but you can also use it for creative purposes.Credit cost might be deceiving, it's 0.8 credits for a 20 step 1024x1024 pixels image, the ordinary SD amount.Main featuresYou can take the style of the image to influence the inpainting, and you can do it in more ways.The smaller the mask, the finer the details (in case of hallucinations try increasing the CFG)Usual inpainting controls included (mask blur, mask expand, denoise value)Cover photo credit Bùi Hoàng Long