Preview for Unified Wan I2V Animation System
Unified Wan I2V Animation System workflow diagram

Run this workflow on InstaSD

Get started in minutes! Run this ComfyUI workflow online - no setup required.

Description

Unleash the full potential of Wan Image-to-Video (I2V) with this ultimate all-in-one animation workflow that merges several beloved effectsβ€”Crush, Rotate 360, Cakeify, Muscle, Inflate, and Squishβ€”into one seamless system. Ideal for creators aiming to generate high-quality character transformation or stylized animation from a single image with minimal setup.


🎯 Features

  • All-in-One Workflow – Combines multiple popular Wan I2V effects (Crush, Rotate, Cakeify, Muscle, Inflate, Squish) in one setup.
  • Real-Time Video Generation – Converts a single input image into a dynamic animation using ComfyUI + VHS.
  • High Customization – Uses flexible positive prompts to easily guide specific transformation styles.
  • CLIP Vision Integration – Ensures visual consistency by using image embedding alongside text prompts.
  • Multi-Effect Control – Swap between effects by switching workflow conditionings or prompt structures.
  • High Quality Output – Powered by wan2.1_i2v_480p_14B_fp8_scaled for optimized animation fidelity.

πŸ’‘ Use Cases

  • Character Animation – Transform a static character into a dynamic animated effect with one click.
  • Stylized FX Shorts – Quickly prototype or produce quirky and fun animated sequences for social media or motion design.
  • VTuber/Influencer Content – Create quick transformations or visual gags for livestream or short-form content.
  • Experimental AI Art – Combine squash-and-stretch, muscle morphs, or surreal inflation for artistic expression.
  • Looped Effects for Video Edits – Use output as overlays or inserts in video montages.

βš™οΈ How It Works

  1. Load the Workflow – Import the .json ComfyUI file corresponding to the combined effect version.
  2. Choose an Input Image – Use the LoadImage node to select your character base.
  3. Set Prompts – Adjust the positive/negative prompt nodes for desired effect (e.g., "inflated body", "crushed pose", "cake texture").
  4. Modify Settings (Optional) – Customize sampler steps, frames, model strength, or add LoRA effects.
  5. Generate Video – Run the flow to generate frames and compile them via the VHS_VideoCombine node.
  6. Preview & Export – Output will be saved as an MP4 in your ComfyUI output folder with loop options.

Credits: Pixaroma

Models

FileDestinationSource
wan2.1_i2v_480p_14B_fp8_scaled.safetensors/ComfyUI/models/unetDownload
squish_18.safetensors/ComfyUI/models/lorasDownload
umt5_xxl_fp8_e4m3fn_scaled.safetensors/ComfyUI/models/clip_visionDownload
wan_2.1_vae.safetensors/ComfyUI/models/vaeDownload
clip_vision_h.safetensors/ComfyUI/models/clip_visionDownload
umt5_xxl_fp8_e4m3fn_scaled.safetensors/ComfyUI/models/text_encodersDownload
inflate_20_epochs.safetensors/ComfyUI/models/lorasDownload
rotate_20_epochs.safetensors/ComfyUI/models/lorasDownload
cakeify_16_epochs.safetensors/ComfyUI/models/lorasDownload
crush_20_epochs.safetensors/ComfyUI/models/lorasDownload
muscle_18_epochs.safetensors/ComfyUI/models/lorasDownload

Nodes

VAEDecodeCLIPVisionEncodeCLIPTextEncodeVHS_VideoCombineUNETLoaderMarkdownNoteWanImageToVideoVAELoaderNoteCLIPVisionLoaderLoadImageCLIPLoaderKSamplerLoraLoaderModelOnly