AudioCraft
AudioCraft (MusicGen, AudioGen, EnCodec)
Type a vibe, get a song
Latest: AudioCraft 1.x with MAGNeT & JASCO additions (2024)
Meta's open generative-audio library (August 2023): MusicGen turns text prompts into music, AudioGen generates environmental sounds, and EnCodec is the neural audio codec underneath. Later gained MAGNeT (non-autoregressive, faster generation) and JASCO (chord/beat-conditioned music). Code is MIT; model weights CC-BY-NC.
Why it matters
The first widely used open text-to-music stack: MusicGen, AudioGen and the EnCodec codec shipped with MIT code and downloadable weights, turning music generation into reproducible research rather than a closed product. The official Hugging Face Space has drawn over 5,000 likes and the library seeded a wave of fine-tunes.
Facts
- ~23,600 GitHub stars — Meta's most popular audio repo.
- MusicGen was trained on 20,000 hours of licensed music.
- EnCodec compresses audio at rates that beat MP3 at 10x smaller bitrates in listening tests.
Try it yourself
Official MusicGen demo (Hugging Face Space) ↗ pip install audiocraft ↗ Model on Hugging Face (musicgen-small) ↗
Sources
arXiv ↗GitHub · audiocraft ↗Meta AI blog ↗