Wan 2.2

Wan 2.2

Packages the Wan 2.2 video models and related files so ComfyUI can load them, covering image to video, text to video, camera control and more. For people who generate video in ComfyUI.

Made by Comfy-Org on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: This repo gathers Wan 2.2 model files in a form ComfyUI can load. It pulls together releases from Wan-AI, alibaba-pai, nvidia and lightx2v, so one download covers image to video, text to video, speech to video, character animation, camera control, inpainting and image editing.

Who it suits: People who make AI video inside ComfyUI and want the model, text encoder, VAE and LoRA files in one place.

Industry | Who uses it | How they use it | Example
Film and media | Video creators | Generate clips from a text prompt or a start image | Turn a still image into a moving shot
Marketing | Content teams | Build short ads with camera control | Add a camera move to a product shot
Animation | Animators | Drive a character or a scene from audio or a control video | Animate a scene with the speech-to-video workflow

If you do not use ComfyUI, you can skip it and use the original model repos.

Where to use it:

• ComfyUI on Windows, macOS or Linux
• The ready-made ComfyUI workflow templates for each task

Use case | Who | What they get back
Image to video | Creators | A video that starts from your image
Text to video | Creators | A video made from a prompt
Image editing | Designers | An edited image from the ChronoEdit workflow

How to set it up:

• ComfyUI installed
• A GPU with enough memory for the 14B or 5B file you choose
• Python with pip, to download the files

1. Install the Hugging Face hub package.
2. Download the repo.
3. Copy each file into the matching folder under ComfyUI models.
4. Open a workflow template and select the files.

pip install -U huggingface_hub
hf download Comfy-Org/Wan-2.2

The card sorts the files into these folders under ComfyUI/models:

• audio_encoders: the wav2vec2 audio encoder
• diffusion_models: the Wan 2.2 and ChronoEdit model files
• loras: the distill and relight LoRAs
• text_encoders: the umt5_xxl text encoders
• vae: the Wan 2.2 and Wan 2.1 VAE files

How to use it: Open one of the workflow templates in ComfyUI, load the matching model files, add your prompt or start image, and run it.

Tool | What it does
Image to Video | Animates a start image with the 14B model
First-Last Frame to Video | Fills in motion between a first and a last frame
Text to Video | Makes video from a prompt with the 14B model
5B Video Generation | Runs the smaller 5B model
Fun Control and Fun Camera Control | Guide the video with a control input or a camera path
S2V | Drives the video from audio
ChronoEdit | Edits an image

Example: You open the 14B image-to-video workflow in ComfyUI, pick the Wan 2.2 files, add a still photo and a short prompt, and run it. A video that starts from your photo comes back.

For newcomers and experts: If you are new, start with one workflow template and the files it names. Get a single clip working before you try the control or camera variants.

Many 14B files come as high-noise and low-noise pairs, in fp16, bf16 and fp8 scaled forms, and the smaller fp8 files need less memory. LoRAs for 4-step sampling from lightx2v are included for faster runs. Pair each model with the VAE and text encoder its workflow expects.

Download the repo files
Place them in ComfyUI model folders
Open a workflow template
Select files and add your prompt
Run to get the video

Good to know:

• These are repackaged files, and the original repos hold the upstream releases
• The file set is large, so pick only what your workflow needs
• License: Apache 2.0
• Cost to run: none named, you supply the GPU
• Data it reaches: your prompts and images, and the files downloaded from Hugging Face
• Not for: use outside ComfyUI, since the files are packaged for it

On Hugging Face, 13 November 2026

• Task: image to video
• Library: diffusion-single-file
• Downloads in the last 30 days: 9,760,687
• Likes: 1,020
• License: apache-2.0
• Based on: nvidia/ChronoEdit-14B-Diffusers
• Quantized versions: 1
• Fine-tunes: 4
• Tags: diffusion-single-file, comfyui, image-to-video

Links

• Model: https://huggingface.co/Comfy-Org/Wan-2.2
• Files: https://huggingface.co/Comfy-Org/Wan-2.2/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download Comfy-Org/Wan-2.2

Then follow the usage in its model card. Nothing in this kit runs by itself.