Skip to content
Help Build a Better ComfyUI Knowledge Base Become a Patron
Tutorial SeriesComfyUI Advanced Tutorialvideowan2.1Wan2.1 LoRA
NEW|

Hi3DGen: A New Framework for High-Fidelity 3D Geometry Generation through Normal Bridging

04/07/2025

Wan2.1 Video LoRA Workflow

In this example, we will use LoRA models in the Wan2.1 Video workflow. For this demonstration, we’ll use the following model:

Access and download the model: Squish Effect (Wan2.1 I2V LoRA) Save the model to the ComfyUI/models/loras directory

In the model description, the LoRA trigger words sq41sh squish effect are mentioned, so we need to add these trigger words to our prompts when using the model.

💡

This workflow is modified from the original Wan2.1 Video Workflow. We only need to add the LoRA model to the existing workflow.

  • We won’t repeat the model download and installation instructions here, focusing only on how to add LoRA models to existing workflows
  • We won’t provide separate workflows for 720P and 480P versions, as this guide is primarily about using LoRA models
  • For custom nodes, please refer to the Wan2.1 Video Workflow guide to install the required custom nodes

Native ComfyUI Wan2.1 Video LoRA Workflow

Native ComfyUI Wan2.1 Video LoRA Workflow

JSON file download

Use the following image as input

LoRA Input Image

Native ComfyUI Wan2.1 Video LoRA Workflow

  1. Make sure the Load image node is using the image we prepared
  2. We’ve added a LoraLoaderModelOnly node to the original image-to-video workflow, loading the Squish Effect (Wan2.1 I2V LoRA) model mentioned earlier
  3. Once all original model loading nodes have correctly loaded the models, click the Run button or use the shortcut Ctrl(cmd) + Enter to generate the video

As you can see from this workflow, the method for loading LoRA models is very similar to most image-type workflows you may have used before.

Kijai’s Wan2.1 Video LoRA Workflow

Kijai's Wan2.1 Video LoRA Workflow

JSON file download

Use the following image as input LoRA Input Image

Kijai's Wan2.1 Video LoRA Workflow

  1. We added a WanVideo Lora Select node, loaded the model we used earlier, and connected it to the WanVideo Model Loader node
  2. Make sure the Load image node is using the image we prepared
  3. Once all original model loading nodes have correctly loaded the models, click the Run button or use the shortcut Ctrl(cmd) + Enter to generate the video

GGUF Version Wan2.1 Video LoRA Workflow

Since the GGUF version is very similar to the native ComfyUI version, we’re only providing the workflow file for the GGUF version

Related Resources

LoRA Model Training

Recommended LoRA Models