All open source
Open SourceVoice & Media22k

AudioCraft

Meta's library for high-quality music and sound generation with MusicGen and AudioGen.

AudioCraft is Meta's PyTorch library for audio generation research, bundling the MusicGen text-to-music model, AudioGen text-to-sound model, and the EnCodec neural audio codec. It provides training and inference code so builders can generate music and sound effects from text prompts or melodies and compress audio efficiently.

Repository

facebookresearch/audiocraft

Language

Python

What you'd build with it

  • Generate royalty-free background music or stems from a text prompt or melody
  • Produce sound effects and ambient audio from text for games or video
  • Use the EnCodec codec to compress or tokenize audio inside an ML pipeline

Tags

music-generationaudio-generationmusicgenencodec