Skip to content
← Models
Open weightsMeta/10 trustVerified
MusicGen Large logo

MusicGen Large

Meta's open text-and-melody conditioned music generation model.

Visit official page →

About

MusicGen Large is Meta's open music generation model, released in June 2023 as part of the AudioCraft research program. It is a single-stage transformer language model built on the EnCodec audio tokenizer that generates music from text descriptions and can optionally be steered with a reference melody.

It was trained on 20,000 hours of licensed music, including stock-library tracks from ShutterStock and Pond5, and it helped establish text-to-music as a mainstream generative category. The code is MIT licensed while the model weights carry CC-BY-NC, a non-commercial license, so commercial music products need a different arrangement.

Specs

Context window

Pricing

Free quota

API access

Weights

Open weights

Released

June 12, 2023

Benchmarks

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

music-generationaudiocreativeresearch

Trust score — 8/10

/10 trustVerified

Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 130; Trending score 80/100 from researcher assessment.

More in the musicgen family

Sources