WorkflowListedExternal

VACE + CausVid Lora! FAST & TOP QUALITY — VACE Pose2Vid

These workflows are only for Pose2Video, so check out my channel for the other VACE Examples, and please consider subscribing if you like this kind of content. It helps me create more!

Sourcecivitai.com/models/1598288

TaskImage to imageFamilyWanVersionEstimated VRAMNot yet characterizedUnratedgrade

Runtime profile

Access: External workflow
Status: Listed
Version: NativeVacePose2Vid.json
Source: Civitai
Updated: Nov 3, 2026

Source description

No release description captured.

What you need

Model files: 24.1 GiB known subtotal

Remaining hardware or download-size estimates are not available yet.

Requirements

Models

dw-ll_ucoco_384_bs5.torchscript.pt
Found
Checkpoint · 129 MB · PT · Unknown
FramePackI2V_HY_bf16.safetensors
Found
Diffusion model · 24.0 GB · SAFETENSORS · Unknown
native_umt5_xxl_fp8_e4m3fn_scaled.safetensors
Not resolved
Text encoder · Unknown
native_wan_2.1_vae.safetensors
Not resolved
VAE · Unknown
Wan21_CausVid_14B_T2V_lora_rank32.safetensors
Not resolved
LoRA · Unknown
yolox_l.onnx
Not resolved
Object Detector · Unknown

Node packs

comfyui_controlnet_aux
Possible
Node pack · Registry
comfyui_layerstyle
Found
Node pack · Registry
comfyui-easy-use
Found
Node pack · Registry
comfyui-kjnodes
Found
Node pack · Registry
comfyui-videohelpersuite
Found
Node pack · Registry

AI-generated commentary

AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.

Purpose

This Wan Video workflow is for Pose2Video generation; its graphs include image and video loading, text-prompt encoding, VACE processing, and a video-combine output stage.

Flow

The archive contains two workflow files: NativeVacePose2Vid.json (16 nodes) and VaceWrapper.json (41 nodes).

Inputs and outputs

The graphs list LoadImage, VHS_LoadVideo, CLIPTextEncode, and VHS_VideoCombine; they do not specify which inputs are required or the output file format.

Setup

VaceWrapper.json’s setup note gives the location ComfyUI>models>diffusion_models and the selector WanVideo VACE Model Select for Wan2_1_VACE_1_3B_preview_bf16.safetensors; it says the other models are the same as base Wan2.1.
The description points to a video for model-download links and a Patreon post for direct links in .sh files.

Dependencies

Both files list comfyui_controlnet_aux, comfyui_layerstyle, comfyui-easy-use, comfyui-kjnodes, and comfyui-videohelpersuite; VaceWrapper.json also lists ComfyUI-WanVideoWrapper, Fast Groups Muter (rgthree), GetNode, and SetNode.
NativeVacePose2Vid.json lists dw-ll_ucoco_384_bs5.torchscript.pt, FramePackI2V_HY_bf16.safetensors, native_umt5_xxl_fp8_e4m3fn_scaled.safetensors, native_wan_2.1_vae.safetensors, Wan21_CausVid_14B_T2V_lora_rank32.safetensors, and yolox_l.onnx.
VaceWrapper.json lists depth_anything_v2_vitl.pth, dw-ll_ucoco_384_bs5.torchscript.pt, native_umt5_xxl_fp8_e4m3fn_scaled.safetensors, Wan2_1_VAE_fp32.safetensors, Wan2_1-VACE_module_14B_bf16.safetensors, wan2.1_t2v_14B_fp8_e4m3fn.safetensors, Wan21_CausVid_14B_T2V_lora_rank32.s Е

Configuration

The description says the examples use VACE 14B and also mentions a 1.3B model; VaceWrapper.json names the 1.3B preview file.

Cautions

Before setup, check which ComfyUI package provides comfyui_controlnet_aux; that name does not identify one specific package.
Suggestion · not verified
Before using VaceWrapper.json, check how to obtain Fast Groups Muter (rgthree), GetNode, and SetNode; the workflow names them but does not name their package owners.
Suggestion · not verified

Does this need to be edited?

Sources

1 source

CivitaiPrimaryVersion 1808645

civitai.com/models/1598288

Verified Nov 3