Wan2.2 TI2V 5B GGUF

Wan2.2 TI2V 5B GGUF

A 5B-parameter text-to-video model converted to GGUF for ComfyUI, about 2.5 GB at 4-bit by estimate, for English and Chinese prompts. It suits ComfyUI users who want video generation on their own computer.

Made by QuantStack 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: Wan2.2 TI2V 5B GGUF is a direct conversion of the Wan-AI Wan2.2-TI2V-5B video model into the GGUF format, made by QuantStack. It has 5B parameters, makes video from text, and takes about 2.5 GB at 4-bit by estimate. Because it is a quantized copy, the original model's license terms and usage limits still apply. It has been downloaded about 495,000 times in the last month.

Who it suits: It suits ComfyUI users who want to generate video clips on their own computer and are happy to add a custom node. The card is short, so some reading of the original model's notes helps.

Industry | Who uses it | How they use it | Example
Creative hobby | ComfyUI users | Generate video clips from written prompts | A short clip of waves at dusk
Media | Visual designers | Sketch moving scenes before a full production | Testing the motion of a product idea
Research | Video model researchers | Compare a quantized copy with the original model | Checking how output differs at 4-bit

People who use other tools than ComfyUI should check first, since the card only names the ComfyUI-GGUF custom node.

Where to use it:

• ComfyUI with the ComfyUI-GGUF custom node by city96
• Windows, macOS or Linux, wherever ComfyUI runs
• Prompts in English or Chinese

Use case | Who | What they get back
Text to video | Hobbyists and designers | A generated clip from a written prompt
Smaller downloads | People with limited disk space | A model file of about 2.5 GB at 4-bit, by estimate
Comparing with the original | Researchers | Output from a quantized copy of Wan2.2-TI2V-5B

How to set it up:

• ComfyUI installed on your computer
• The ComfyUI-GGUF custom node, installed by following its own readme
• Disk space for the model file, about 2.5 GB at 4-bit by estimate

1. Install the Hugging Face hub package.
2. Download the model files.
3. Install the ComfyUI-GGUF custom node by following its readme.
4. Put the model file in the unet folder under ComfyUI models.

pip install -U huggingface_hub

hf download QuantStack/Wan2.2-TI2V-5B-GGUF

How to use it: Build a video workflow in ComfyUI and load the model with the loader node that the custom node adds. The card gives no workflow or settings, so take those from the notes for the original Wan2.2 model.

Tool | What it does
ComfyUI-GGUF | A custom node that lets ComfyUI load GGUF model files
unet folder | The place under ComfyUI models where the model file goes
hf download | Fetches the repository files to your computer

Example: You download the file, put it in the unet folder, and open a text-to-video workflow in ComfyUI. You write a prompt in English or Chinese and queue it. The model returns a short generated clip, which you can compare with the original model.

For newcomers and experts: A quantized model stores its numbers in fewer bits, so the file is smaller, at some cost to accuracy. A good first step is to read the original Wan2.2-TI2V-5B card, since this repository only holds the converted file.

The name TI2V suggests the original can also start from an image, though this card lists only text to video. The card does not say which quantization levels the repository holds.

You write a text prompt
ComfyUI-GGUF loads the GGUF file
The model generates the frames
A video clip comes out

Good to know:

• License: Apache 2.0, and the card says the original terms and usage limits stay in effect.
• Cost to run: none named. You need your own computer and no paid service.
• Data it reaches: it downloads files from the Hugging Face Hub, then generates video on your machine.
• Not for: work that needs documented settings or quality results, since the card gives none.

On Hugging Face, 13 November 2026

• Task: text to video
• Library: gguf
• Parameters: 5B
• Weights: about 2.5 GB at 4-bit (estimate)
• Formats: GGUF
• Downloads in the last 30 days: 480,229
• Likes: 259
• License: apache-2.0
• Languages: en, zh
• Based on: Wan-AI/Wan2.2-TI2V-5B
• Tags: gguf, ti2v, text-to-video

Links

• Model: https://huggingface.co/QuantStack/Wan2.2-TI2V-5B-GGUF
• Files: https://huggingface.co/QuantStack/Wan2.2-TI2V-5B-GGUF/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download QuantStack/Wan2.2-TI2V-5B-GGUF

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