Wan2.1-T2V-1.3B

Wan2.1-T2V-1.3B

Makes short 480P videos from a text prompt, small enough for most consumer GPUs with about 8 GB of video memory. For creators and researchers who want local text-to-video.

Made by Wan-AI 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: Wan2.1-T2V-1.3B is the small text-to-video model of the Wan2.1 video family from Wan-AI. It turns a written prompt into a 480P video and can also render Chinese and English text inside the picture. The record lists 1.4B parameters.

Who it suits: Creators, researchers and small teams who want to generate video on their own GPU instead of a paid service.

Industry | Who uses it | How they use it | Example
Marketing | Content teams | Draft short video concepts from a prompt | A five-second product mood clip
Film and video | Independent filmmakers | Test shots before filming | A rough clip of a street scene
Research | Academic labs | Study video generation on modest hardware | A baseline for experiments
Education | Creative classes | Teach text-to-video prompting | Students compare prompts and clips

You can skip it if you need 720P output or image-to-video. The card points to the larger 14B models for 720P.

Where to use it:

• Your own computer with a consumer GPU, using the model's code repository
• Windows, macOS or Linux, with Python for the download step
• The Gradio demo that the card lists as available

Use case | Who | What they get back
Text to video | Creators | A 480P clip from your prompt
Text in video | Designers | Chinese or English words rendered in the clip
Low-memory runs | Hobbyists | Generation within about 8.19 GB of video memory

How to set it up:

• Python and pip on your computer
• A consumer GPU, since the card says the model needs only 8.19 GB of VRAM
• A recent PyTorch, at least version 2.4.0

1. Clone the Wan2.1 code repository, using the git clone line on the card.
2. Install its requirements.
3. Download this model's files with the hf command.

cd Wan2.1

# Ensure torch >= 2.4.0
pip install -r requirements.txt

pip install -U huggingface_hub
hf download Wan-AI/Wan2.1-T2V-1.3B

How to use it: Run the generation script from the Wan2.1 repository, as its README describes, and point it at the downloaded files. You write a prompt and get a video file back. The card says a 5-second 480P clip takes about 4 minutes on an RTX 4090 without extra optimization.

hf download Wan-AI/Wan2.1-T2V-1.3B

Example: You write a prompt describing a cat walking along a beach at sunset. After a few minutes of generation on a capable GPU, you get a 5-second 480P clip of that scene.

For newcomers and experts: If you are new, check that your GPU has at least 8 GB of video memory, then generate one short clip to see what to expect. Expect minutes, not seconds.

If you know the field, the card says the family uses a video VAE called Wan-VAE that encodes and decodes 1080P video of any length while keeping temporal information. This 1.3B text-to-video checkpoint targets 480P. Larger 14B checkpoints support 480P and 720P.

Text prompt
Wan2.1 T2V-1.3B diffusion model
Wan-VAE decoder
480P video file

Good to know:

• License: Apache 2.0
• Cost to run: your own GPU hardware, with about 8.19 GB of VRAM per the card
• Data it reaches: the files you download and the prompts you write
• Not for: 720P output, which the card assigns to the 14B models
• Versions: others have published 10 quantized and 130 fine-tuned copies

On Hugging Face, 13 November 2026

• Task: text to video
• Library: diffusers
• Parameters: 1.4B
• Weights: about 5.7 GB in F32; about 709 MB at 4-bit (estimate)
• Formats: Safetensors
• Downloads in the last 30 days: 91,394
• Likes: 676
• License: apache-2.0
• Languages: en, zh
• Quantized versions: 11
• Fine-tunes: 129
• Tags: diffusers, safetensors, t2v, video, video-generation, text-to-video

Links

• Model: https://huggingface.co/Wan-AI/Wan2.1-T2V-1.3B
• Files: https://huggingface.co/Wan-AI/Wan2.1-T2V-1.3B/tree/main

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download Wan-AI/Wan2.1-T2V-1.3B

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