A ComfyUI workflow pushes Wan 2.2 image-to-video out to roughly 45 seconds

embryo10 · reddit · 2026-08-04

A detailed workflow post shows how to build effectively unbounded Wan 2.2 image-to-video generation in ComfyUI.

The workflow has a 5-second initial block plus up to eight optional 5-second extension blocks, reaching about 45 seconds before frame loss becomes an issue.
It can start from either a still image or an existing video.
The initial block supports start-frame and end-frame guidance for image-to-video.
Each block has its own prompt selector, duration control, fixed seed, and independent LoRA section.
The author recommends GGUF loaders for low-VRAM systems and notes that the latest ComfyUI frontend versions may break subgraphs; they suggest pinning comfyui-frontend-package==1.39.2.
The post also lists changelog notes from earlier versions, including video extension support and better compatibility fixes.

More from coding & agent