# openai/jukebox

Code for the paper "Jukebox: A Generative Model for Music"

Repository: https://github.com/openai/jukebox
Canonical: https://ross.abutalabs.com/products/jukebox
Homepage: https://openai.com/blog/jukebox/
Language: Python
License: NOASSERTION
License Family: other
Topics: paper, audio, music, pytorch, generative-model, vq-vae, transformer
Archived: true
Last push: 2024-06-19T05:14:24+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2317, "days_push": 805, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 8029, forks 1443 (observed 2026-08-28T04:10:11.948601+00:00)

## What it is
OpenAI's research code and pretrained model weights for Jukebox, a generative neural network that produces music with vocals as raw audio from genre, artist, and lyrics conditioning. It implements a VQ-VAE plus transformer priors in PyTorch and is archived with no further updates.

## Use cases
- generate music with singing from lyrics and genre
- sample raw audio songs with a neural network
- research generative models for music
- experiment with VQ-VAE audio compression
- reproduce the Jukebox paper results
- fine-tune a music generation model on my own data

## When to choose
- you need raw-audio music generation with vocals and have large GPU resources
- you are doing research on hierarchical audio generative models
- you want to study or extend the Jukebox architecture

## When to avoid
- you need a maintained library or timely bug fixes
- you lack high-end GPUs - sampling is extremely slow (hours for seconds of audio)
- you need fast, production-ready music generation
- you need a permissive license - the license is non-standard

## Facets
- artifact type: library
- maturity: abandoned
- function: deep-learning, audio-processing, machine-learning, llm-training
- domain: deep-learning, machine-learning, artificial-intelligence
- platform: python
- tags: music-generation, generative-model, vq-vae, transformer, pytorch, research-code, raw-audio, archived, audio, linux, gpu

## Member repositories
- openai/jukebox (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:11.948601+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:31:48.256079+00:00, confidence not recorded.
  - readme: https://github.com/openai/jukebox (fetched 2026-08-28T04:10:11.948601+00:00, sha 026c168bcedd)
  - homepage: https://openai.com/blog/jukebox/ (fetched 2026-08-29T08:29:16.791261+00:00, sha af24ce660b30)
  - site_page: https://openai.com/about (fetched 2026-08-29T08:29:16.801068+00:00, sha de4b627b33e3)
  - site_page: https://developers.openai.com/api/docs (fetched 2026-08-29T08:29:16.802995+00:00, sha d617c5215f2a)
  - site_page: https://developers.openai.com/ (fetched 2026-08-29T08:29:16.804973+00:00, sha 7d240c906018)
- Data as of 2026-08-30T08:39:29.467469+00:00.