A simple web user interface built on top of the MMAudio library, enabling both video-to-audio and text-to-audio generation through an interactive browser interface. Designed for easy local deployment and experimentation in Python environments.
Key Features
Video‑to‑audio conversion using MMAudio
Text‑to‑audio generation via MMAudio library
Minimalistic browser-based UI built with Gradio or Flask
Automatic download of required model weights (MMAudio, CLIP, Motion encoder, BigVGAN vocoder)
Quick startup through a single script
Python 3 environment ready, including virtual environment support