Pose-controlled character art with ComfyUI ControlNet

A ComfyUI workflow that renders a character image matching a reference pose skeleton.

AI image generation · 30-45 min · Updated 2026-06-01

First win

A generated character appears in the pose extracted from the reference image.

Best for

ComfyUI users with a working SDXL workflow

Needs

ComfyUI installed, SDXL Base model, Pose reference image

Example outcome

A pose reference becomes a posed character render.

This preview shows the ControlNet pipe: an OpenPose preprocessor extracts a skeleton from the reference and a generated character follows it.