ModelScope Text To Video

ModelScope Text To Video

Turns an English text prompt into a short 256x256 video clip with the ModelScope text-to-video diffusion model, in a Gradio demo. For researchers and curious makers trying text-to-video generation.

Made by shamedrank10 on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This Space is a Gradio demo of ModelScope Text2Video, also called VideoFusion, a text-to-video diffusion model of about 1.7 billion parameters from Alibaba DAMO and ModelScope. You type an English prompt and it returns a short video at 256x256 pixels. It runs on ZeroGPU at Hugging Face.

Who it suits: Researchers, students and hobby makers who want to see what an early open text-to-video model can do, with tuning controls close at hand.

Industry | Who uses it | How they use it | Example
Research | Video generation researchers | Check the model's output before using it in a study | Generate a clip and note frame detail and motion
Education | Students of diffusion models | See how text becomes video | Change the seed and compare two clips of one prompt
Creative | Hobby makers | Make short silent clips from an idea | Type a short scene and press Generate video
Product teams | Prototype builders | Judge whether text-to-video fits an idea | Try the example prompts the page offers

If you need commercial output or readable text inside the video, this model is not the right choice.

Where to use it: It runs in two places.

• A web browser, on the Space's page at Hugging Face with nothing to install
• A local copy on Windows, macOS or Linux, which needs Python 3.12

Use case | Who | What they get back
Make a clip | Makers | A short MP4 video from an English prompt
Tune a run | Researchers | Different clips from changed frames, steps, fps or seed
Steer away from things | Makers | A clip made with a negative prompt applied

How to set it up: Using it online needs nothing. A local copy needs a few things.

• Python 3.12 and pip
• The packages listed in requirements.txt
• A GPU suited to the model, since the page was built for ZeroGPU

1. Install the Hugging Face tools.
2. Download the Space into a folder.
3. Move into that folder.
4. Install Gradio and the packages from requirements.txt.
5. Start the app.

pip install -U huggingface_hub
hf download shamedrank10/modelscope-text-to-video --repo-type space --local-dir modelscope-text-to-video
cd modelscope-text-to-video
pip install gradio==6.30.0
pip install -r requirements.txt
python app.py

How to use it: Write your scene in English, or choose a ready-made example prompt. Open Advanced options if you want to change the frame count, inference steps, fps, seed or negative prompt. Then press Generate video and wait for the clip.

Tool | What it does
Prompt box | Takes your English description of the scene
Advanced options | A panel with controls for frames, steps, fps, seed and what to avoid
Generate video | Starts the run and returns the clip

Example: You type "a cat walking along a beach" and press Generate video. After a short wait the page returns a 256x256 clip. With 16 frames at 8 fps, it lasts about two seconds.

For newcomers and experts: Start with an example prompt and the default settings. Short, concrete scenes work best for a first try, and the model reads English only.

The demo loads an fp16 pipeline once at startup through the diffusers port of the model. It exports MP4 files with imageio-ffmpeg. Time grows with frame count and inference steps. The model cannot draw clear text and may repeat biases from its training data.

You type an English prompt
The fp16 pipeline runs on ZeroGPU
The model denoises frames into a video
The page exports and shows an MP4

Good to know: Output size is fixed at 256x256, so clips are small and short.

• License: CC BY-NC 4.0, for non-commercial use only
• Cost to run: ZeroGPU time at Hugging Face, with no key named
• Data it reaches: the prompts you type into the page
• Not for: commercial use, non-English prompts or videos with clear text

On Hugging Face, 13 November 2026

• Likes: 1
• Built with: gradio
• License: cc-by-nc-4.0

Links

• Space: https://huggingface.co/spaces/shamedrank10/modelscope-text-to-video
• Files: https://huggingface.co/spaces/shamedrank10/modelscope-text-to-video/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.