ComfyUI_StoryDiffusion

Why we included this project

ComfyUI users who need a character's face to stay consistent across a sequence of generated images will get the most out of this node collection. It bundles StoryDiffusion and a wide range of identity-preservation backends like PhotoMaker, PuLID, IP-Adapter, and InstantCharacter, so the same look carries from one panel to the next without hand-wiring each method into your graph. The workflow nodes slot into an existing ComfyUI setup, with support for SDXL checkpoints, image-to-image via PhotoMaker, and the newer Qwen-Image editing models. That covers concept artists, comic makers, and anyone running character-consistent generation pipelines. One practical note: the identity-heavy modes need the insightface library installed before they will run.