Wan2.2 FunControl FreeLength V2V with Background Changer (Version 2) — v2.0

This workflow creates videos from reference images, with background control and prompt-based pose approximation. You provide reference images, pose prompts, and main-screen settings, and it outputs a video.

Set generation time and processing-unit time separately, then sample for the calculated number of loop iterations.

The description says video resizing occurs for each loop and reduces the risk of running out of memory.

Enter reference images, prompts to approximate the reference pose, and the parameters shown on the main settings screen.

The description says it can create videos of any length, subject to your machine specifications.

The listed base model is Wan Video 2.2 I2V-A14B.

This workflow is version v2.0, and the listed publisher is Usako USA.

Listed node packs include: comfyui_essentials; comfyui-easy-use; comfyui-custom-scripts; comfyui-frame-interpolation; comfyui-kjnodes; comfyui_controlnet_aux; comfyui-ollama; masquerade-nodes-comfyui; comfyui-gguf.

Additional listed node packs are: comfyui_layerstyle; comfyui-rmbg; ComfyUI_LayerStyle_Advance; ComfyUI-WanVideoWrapper; was-ns; derfuu_comfyui_moddednodes; rgthree-comfy; comfyui-videohelpersuite; ComfyUI-tbox.

Listed model files include: dw-ll_ucoco_384_bs5.torchscript.pt; alimama-creative_FLUX.1-Turbo-Alpha.safetensors; t5xxl_fp8_e4m3fn.safetensors; umt5_xxl_fp8_e4m3fn_scaled.safetensors; clip_l.safetensors; flux1-kontext-dev-Q6_K.gguf; rife49.pth; Wan2.2-Fun-A14B-Control_HighNoise-Q6

Most settings can be completed on the main settings screen; detailed settings are available on yellow nodes.

For pose control, use prompts to approximate the reference pose; the description says Flux Kontext plus ControlNet currently produces errors, and the cause is not given.

Suggestion · non vérifié

Before using it, verify that each listed node pack and model file is available in your setup.

Suggestion · non vérifié