📟Wan2.2 I2V/FFLF Small Workflow — 🟢v1.0

This ComfyUI workflow uses image input to make Wan Video 2.2 image-to-video clips, with first/last-frame variants, GGUF or native Safetensor versions, and video post-processing.

Subgraphs group parts of the workflow into a compact layout.

A separate Painter-node version adds motion amplitude.

Frame options include interpolation, model-based upscaling, trimming, and sharpening; you can also save the last frame or select frames.

To open the workflow embedded in the posted video, download the video and drag it into ComfyUI; ComfyUI-VideoHelpSuite is needed.

The included guide covers ComfyUI-GGUF installation on desktop and portable setups, plus ComfyUI-PainterFLF2V for the Painter version.

Listed node packs include ComfyUI-Easy-Use, ComfyUI-KJNodes, ComfyUI-Frame-Interpolation, ComfyUI-VideoHelperSuite, ComfyUI_essentials, and rgthree-comfy.

The post-processing workflow lists RealESRGAN_x2plus.pth and rife47.pth.

ComfyUI-GGUF is listed for the GGUF version, and ComfyUI-PainterFLF2V for the Painter version.

Sage Attention and Lora Stackers are included; WanNAG is described as strengthening the negative prompt when CFG is 1.

FPS can be set automatically based on frame interpolation and speed, or adjusted manually; manual FPS overrides the automatic setting and speed adjustment.

The creator says the Painter version has had little testing, so consider it experimental.

Vorschlag · nicht geprüft

Stop ComfyUI before installing custom nodes. The creator notes ComfyUI-GGUF may not appear in the newer Manage Extension manager.