Wan 2.2 Video-to-Video Tutorial with VACE 2.2! (Beginner Friendly) — v1.0

This workflow is for Wan 2.2 video-to-video processing: its graph accepts an image, a video path, text conditioning, and VACE start-to-end frame settings, then decodes and combines the result as video.

Its 40-node graph loads image and video media, encodes VACE and text inputs, samples with Wan Video, decodes the result, and combines the video.

Inputs include an image, a video path, cached text encoding, and VACE start-to-end frame selection.

WanVideoDecode and VHS_VideoCombine provide the graph path for video output.

Find it on Civitai at https://civitai.com/models/2390511, where it is listed by Lordcaocao2025.

The archive is named wan22VideoToVideoTutorial_v10.zip.

Listed node packs: comfyui_controlnet_aux, comfyui-kjnodes, comfyui-various, comfyui-videohelpersuite, and ComfyUI-WanVideoWrapper.

Listed model files: lightx2v_T2V_14B_cfg_step_distill_v2_lora_rank64_bf16.safetensors, wan_2.1_vae.safetensors, Wan2_2_Fun_VACE_module_A14B_HIGH_Q8_0.gguf, Wan2_2_Fun_VACE_module_A14B_LOW_Q8_0.gguf, Wan2.2-T2V-A14B-HighNoise-Q5_K_S.gguf, and Wan2.2-T2V-A14B-LowNoise-Q5_K_S.gguf.

The base model is Wan Video 2.2 T2V-A14B.

Available controls include block swapping, LoRA selection, Torch compile settings, CFG schedules, and VACE start-to-end frames.

Before using it, verify that every listed model file and node pack is available and that comfyui_controlnet_aux refers to the node pack you intend to install.

Suggestion · non vérifié

The source page has no separate description or summary, so review the workflow details before deciding whether it fits your use.

Suggestion