>>18121
>Share the prompt?
I can't.
I mean, I'd want to, but this isn't a "click-to-generate" work. I'm:
- taking 3D render as source;
- switching style to drawn (to avoid photo-reality);
- starting to ask the model (Qwen-Image-Edit) to change some details (add / enhance body fluids, change color, lighting, tune faces and bodies, etc), receiving dozens of versions, choosing few good;
- combining them in graphic editor.
- sometimes doing final pass to remove seams.

So, this isn't really AI work, it involves 3D, AI and 2D steps. One image could consume hours to do, depends on plot and luck.

For switch styling prompt only you may see the author and my comments to him here:
>>16701