Position characters with 3D poses

Place and pose the characters in a 3D view first, then render that into the picture.

You place and pose stand-in figures in a 3D view -- who stands where, who faces whom -- and that arrangement guides the image. It gives you control over blocking that a written description cannot.

Use it when where people stand matters, such as a fight or a crowded room.

Recipes that carry it out: none yet. Browse AI models is the way to make one: install models, build a workflow in ComfyUI, then make a recipe around it. Read more in Techniques.