Free ComfyUI Workflow: Animate Images with Video Pose Transfer & Character Replacement

Bring your static characters to life with unprecedented ease using our Free ComfyUI Wan Animate Workflow. This powerful template allows you to transfer motion from any video onto a still image, creating fluid, dynamic animations. Even more, it includes an advanced mode to completely replace a character in a video with your own.

Whether you're an AI artist, animator, or content creator, this workflow provides a fast and intuitive way to produce complex character animations directly within ComfyUI.

What is the Wan Animate Workflow?The Wan Animate workflow is a comprehensive ComfyUI template built around the innovative Wan Animate model. It leverages the DWPose Estimator to analyze the motion in a source video and applies that exact pose data to a target image frame by frame.

This template comes with two powerful, pre-configured modes:

1. Pose Transfer: Animate a static character image using a reference video as the motion source.
2. Character Replace: Swap a character in a source video with a new character from an image, using intelligent masking to integrate them into the scene.## Key Features

• 🕺 Accurate Pose Transfer: Uses DWPose to precisely copy body, hand, and facial movements from a video to your image.
• 🎭 Seamless Character Replacement: Intelligently masks and replaces characters in existing video footage.
• ⚡ Optimized for Speed: Utilizes the FP8 version of the Animate model for faster generation compared to standard methods.
• 🎨 Lora & Lighting Ready: Comes pre-configured with a Power Lora Loader and lighting LoRAs (Lightx2v and Wan Relight) to enhance your final output.* 🔧 Fully Customizable: Easily adjust FPS, video length, sampling seeds, and masking parameters to fit your project's needs.

How to Use: Basic Pose Transfer

Follow these steps to animate your character from a reference video.

1. Load Your Image and Models

Start by loading your character image into the "Load Image" node. The workflow is pre-set to handle model loading automatically, including the Animate model, VAE, and essential LoRAs.

2. Upload Your Reference Video

In the "Load Video (upload)" node, select the video file containing the motion you want to copy. You can adjust settings here:

• FPS: 16 FPS is a good starting point for smooth results.
• Frame Cap: Limit the number of frames to control the output video length.

The workflow will automatically process the video through the DWPose Estimator to create a pose map, which you can preview.

3. Write Your Prompt

Add a positive prompt describing your character and scene. The negative prompt is pre-filled, but you can adjust it if needed.

4. Generate Your Video

Click "Queue Prompt" to start the generation. The KSampler node will use the pose data to animate your image. The final output is a high-quality video with your character performing the actions from the reference footage.

How to Use: Character Replace Workflow

This advanced mode lets you replace a person in a video.

1. Set Up the MaskThe "Replace" workflow adds a masking section.

1. Run the workflow once. It will process the video's pose and then pause, loading the first frame into the "Points Editor" node.
2. Stop the queue.
3. In the Points Editor, use Shift + Left Click to add green points to the character you want to replace.
4. Use Shift + Right Click to add red points to the background or areas you want to exclude.
5. You can adjust the growmask and blockify_mask parameters to refine the mask's edges and shape.

2. Generate the Final Video

Once your mask is set, click "Queue Prompt" again. The workflow will use the mask to isolate the character and seamlessly insert your new character from the loaded image, applying the original motion.

Who is This Workflow For?

• AI Animators: Create complex character animations without keyframing.
• Content Creators: Quickly produce engaging video content and memes.
• Game Developers: Prototype character movements and animations.
• AI Enthusiasts: Explore the cutting edge of AI-driven video and motion transfer technology.

Frequently Asked Questions (FAQ)

What is DWPose? DWPose is an advanced pose estimation model that accurately detects the position of human joints, hands, and facial features. This workflow uses it to create the motion data needed for animation.

How long does it take to generate a video? The first run for a new video may take a few minutes as the DWPose Estimator processes the entire clip. Subsequent runs are much faster, as the pose data is cached.

Can I use my own LoRAs? Yes! The workflow includes a Power Lora Loader, making it easy to add multiple character, style, or concept LoRAs to influence the final output.

Download the Wan Animate Workflow for ComfyUI

Ready to start animating? Download this simple, fast, and powerful workflow to unlock incredible pose transfer and character replacement capabilities in ComfyUI.

[Download the Wan Animate Workflow Now]