It's crazy how quickly AI image models have become more flexible. I remember prompting anything involving hands or feet or any pose more complicated than standing or sitting was guaranteed to make an anatomical clusterfuck, but now it can do upside-down,handjob, foot focus prompts pretty consistently.
https://files.catbox.moe/crhdhi.pnghttps://files.catbox.moe/i3z2iq.pnghttps://files.catbox.moe/47q6ii.pnghttps://files.catbox.moe/6bp6du.png