ForkFree

Video and audio

mochi

by genmoai

Mochi 1 is an open video generation model from Genmo that turns text prompts into short video clips. It comes with a web interface, a command-line tool, a Python API and a trainer for customising the model on your own videos.

Worth a lookEstablished, with a few things to check (listed below).

View on GitHub

Stars
3.8k
Licence
Apache-2.0
Runs on
A gpu machine, with a GPU
Setup
Technical
Language
Python
Age
781 days

What you can do with it

  • Generate short 480p video clips from written prompts using a web interface or the command line
  • Fine-tune the model with LoRA on your own videos to capture a particular style or subject
  • Call the model from Python code to build video generation into your own apps

Make money with it

A freelancer could sell short AI-generated video clips or branded animated backgrounds to small businesses for adverts and social media.

An idea, not financial advice. Check the licence before charging for anything built on it.

Ways a business could use it

  • Produce short stock-style video clips or social media visuals without paying per-clip fees to a hosted video service
  • Offer custom-styled video generation to clients by fine-tuning a LoRA on their brand footage

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

Clone the repository, install with uv in a Python 3.11 environment, install FFmpeg, then download the model weights with the supplied script.

Follow the official instructions in the project README.

Things to check

  • Needs about 60GB of VRAM on a single GPU (an H100 is recommended), so it is costly to run
  • Research preview with known limitations, such as 480p output and artefacts with extreme motion
  • README says GPU inference and fine-tuning still need qualification before deployment
  • Authors advise extra safety checks before commercial use because of training-data biases and possible NSFW content

Who's talking about it

Checked 01 Nov 2026. Summary drafted by AI from the project's README.

More in Video and audio

VoiceStudio debpalash
Worth a look

VoiceStudio is a desktop, mobile and Docker app that runs speech tools on your own device. It clones and designs voices, dubs videos, transcribes and dictates speech, and turns text into audiobooks in many languages.

Could replace ElevenLabs

94.8k stars +1527 a day AGPL-3.0
hyperframes heygen-com
Worth a look

HyperFrames is an open-source framework that turns HTML, CSS, media and animations into MP4 videos. It is designed so AI coding agents, or people using a command line, can write a page and render it as a video.

Could replace Remotion

76.7k stars +632 a day Apache-2.0
lemo-opuscar lemomo-ai
New and promising

Lemo-Opuscar is a tool for creating films using code, offering 59 pre-made styles or the ability to create your own. It is designed for use with Claude Opus 5.5 and can be installed as a plugin or cloned from GitHub.

Could replace Adobe Spark Video

10.2k stars +364 a day MIT