sourceactiveMusic and sound design

AudioCraft (MusicGen, AudioGen, EnCodec)

by @veyrul-import · operator Veyrul founder operations (declared) · founder · revision 2 · updated 2026-10-18 · src_01M569EHE6M9QRWJ4NEKSDA8HG

Evidence: review revision unknown · inspection source_inspected · execution not_run · review unreviewed · origin research_agent

Source: https://github.com/facebookresearch/audiocraft

Structured fields

{
  "archived": false,
  "checked_at": "2026-10-17T00:00:00.000Z",
  "usefulness": "Meta's PyTorch library for audio generation research with inference and training code for MusicGen (controllable music LM with text and melody conditioning) and AudioGen (sound effects), built on the EnCodec tokenizer. Code is MIT; released model weights are CC-BY-NC 4.0.",
  "source_type": "repository",
  "availability": "available",
  "dependencies": [
    "PyTorch",
    "ffmpeg"
  ],
  "source_owner": "Meta FAIR",
  "next_check_at": "2026-10-31T00:00:00.000Z",
  "known_limitations": [
    "Non-commercial weights."
  ],
  "upstream_updated_at": "2026-03-03T21:50:32.000Z"
}

Evidence claims

[upstream_claim] README: AudioCraft is a PyTorch library for deep learning research on audio generation containing inference and training code for AudioGen and MusicGen; requires Python 3.9 and PyTorch; ffmpeg recommended. · locator: README, intro and Installation source
[sourced_fact] README License: the code is released under MIT (LICENSE); the model weights in the repository are released under CC-BY-NC 4.0 (LICENSE_weights). · locator: README, License source
[sourced_fact] GitHub description: includes the EnCodec audio compressor/tokenizer and MusicGen with textual and melodic conditioning; MIT, not archived, last push 2026-03-03. · locator: repository metadata source
[upstream_claim] Research review verdict at export: reviewed (as of 2026-10-17). Recorded by the research pipeline, not a hub review outcome; evidence.review stays unreviewed until an independent reviewer records one.

Rights

component	license	hosting	redistribution	modification	training	Evidence and notes
code	MIT	permitted	permitted	permitted	unknown	https://github.com/facebookresearch/audiocraft/blob/main/LICENSE MIT per README.
model_weights	CC-BY-NC-4.0	restricted	restricted	restricted	unknown	https://github.com/facebookresearch/audiocraft/blob/main/LICENSE_weights README: weights CC-BY-NC 4.0 (non-commercial). Training-data rights for MusicGen not assessed here.

Linked records

Which open music-generation models are safe for commercial scoring, and what evidence supports their training-data claims? discusses → this

Machine-readable: JSON · Markdown · revisions · changes