CogVideo
by zai-org
CogVideo and CogVideoX are open-source AI models that create short videos from a text description or a starting image. The repository includes code to run the models, fine-tune them on your own footage, and try them in a browser demo.
- Stars
- 13.2k
- Licence
- Apache-2.0
- Runs on
- A gpu machine, with a GPU
- Setup
- Technical
- Language
- Python
- Age
- 1617 days
What you can do with it
- Generate short video clips from a written prompt
- Animate a still image into a video using the image-to-video model
- Fine-tune the model with LoRA on your own videos to get a particular style
Make money with it
You could offer short AI-generated video clips or animated product images as a paid service to local businesses and social media managers.
An idea, not financial advice. Check the licence before charging for anything built on it.
Ways a business could use it
- Produce short social media clips, product teasers or background footage without paying for stock video or a per-clip video service
- Animate product photos or illustrations into short promotional videos for adverts and web pages
Can you use it commercially?
Free to use, change and sell, including in paid products. Keep the copyright notice.
Not legal advice. Read the project's licence file before building a product on it.
Getting it running
Use Python 3.10 to 3.12, install the dependencies with pip install -r requirements.txt, then run the Diffusers inference script or the SAT demo.
Follow the official instructions in the project README.
Things to check
- Needs a capable GPU; the 5B model runs on a desktop card such as an RTX 3060 but slowly
- Prompts work best when rewritten by a large language model, which may need a paid API key
- Output is short and quality varies, so it is not a replacement for commercial services
- Model weights have their own licence terms, so check them before commercial use
Who's talking about it
Checked 01 Nov 2026. Summary drafted by AI from the project's README.
