Skip to content
Local AI Wizard
Your machine
Theme

MusicGen Small

Meta FAIR · facebook/musicgen-small

Weights license: Publisher terms: CC-BY-NC 4.0 weights / MIT code · code and components may use other terms

Rating · Music Arena (instrumental)794Published result · reviewed 2026-10-20
Published repository size5.8 GBAll published files · not runtime memory
Parameters0.3 Bmanufacturer model card
Published components2Listed in the model card

Quality

794 · Rating · Music Arena (instrumental) · the only one of our models on this board, so not ranked here · Source: gclef-cmu/music-arena leaderboard snapshot 2026-07-31 @ 8b330f4 · collected 2026-10-20 · MIT · CC BY 4.0 · models without a result are not drawn (no result ≠ 0)

Rating · Music Arena (instrumental)

Published result · collected 2026-10-20 · gclef-cmu/music-arena leaderboard snapshot 2026-07-31 @ 8b330f4

95% interval
759–829
Votes
463
Board position
14 of 15 (all models, open and proprietary)
Benchmark conditions2026-10-20

Blind pairwise human votes on music-arena.org between clips generated from the same prompt; Bradley-Terry rating with a 95% bootstrap interval, models with 30+ votes. Not a local hardware measurement.

Music Arena instrumental board, votes 2025-07-28 to 2026-07-31 · gclef-cmu/music-arena leaderboard snapshot 2026-07-31 @ 8b330f4

Reported by the publisher

3 results · Meta FAIR's model card · read 2026-10-25
Frechet Audio Distance
4.88
KLD
1.42
Text Consistency
0.27

Values as printed on the card, with the publisher's own settings (card section: Evaluation results); column “facebook/musicgen-small”. Not comparable across publishers and not used in any ranking. Model card

Quality × memory

No comparable quality × memory pair

Image, video and audio models have no shared quality scale, and their memory here is a 16-bit planning estimate, not a documented minimum, so quality is not charted against memory.

Runs on

Enough memory for the estimate 119

The 16-bit planning estimate fits this memory. It is not a reviewed minimum.

Show all 119

The catalog has not reviewed this checkpoint's hardware requirement, so no publisher figure is shown here. Hardware profiles are grouped by a labelled 16-bit planning estimate from the published parameter count (2 bytes per parameter, 20% allowance, 2 GB reserve). Precision, resolution, duration, text encoders and offload change the real requirement. Not a measured run.

Plan memory for this model

Consider it for

Consider MusicGen Small from Meta FAIR for the text to music task (0.3B parameters).

Check first

This catalog has not verified hardware requirements for this entry. Actual memory depends on precision, resolution, duration and offload; consult the official source.

Measured speed

No published measurement for this checkpoint

Speed (real-time factor, or seconds per clip) is shown only when a source publishes it with hardware, build and setup. It is never estimated.

How to run

Hardware requirement not reviewed. This catalog has not verified hardware requirements for this entry. Actual memory depends on precision, resolution, duration and offload; consult the official source.

Documented ways to run it

Copied as printed in the linked source (read 2026-10-18); not run by this site.

Setup stated first by the source · Transformers
pip install --upgrade pip
pip install --upgrade transformers scipy
Transformers · model card · 🤗 Transformers Usage
from transformers import pipeline
import scipy

synthesiser = pipeline("text-to-audio", "facebook/musicgen-small")

music = synthesiser("lo-fi music with a soothing melody", forward_params={"do_sample": True})

scipy.io.wavfile.write("musicgen_out.wav", rate=music["sampling_rate"], data=music["audio"])
Setup stated first by the source · AudioCraft
pip install git+https://github.com/facebookresearch/audiocraft.git
AudioCraft · model card · Audiocraft Usage
from audiocraft.models import MusicGen
from audiocraft.data.audio import audio_write

model = MusicGen.get_pretrained("small")
model.set_generation_params(duration=8)  # generate 8 seconds.

descriptions = ["happy rock", "energetic EDM"]

wav = model.generate(descriptions)  # generates 2 samples.

for idx, one_wav in enumerate(wav):
    # Will save under {idx}.wav, with loudness normalization at -14 db LUFS.
    audio_write(f'{idx}', one_wav.cpu(), model.sample_rate, strategy="loudness")
Setup stated first by the source · Transformers
pip install --upgrade pip
pip install --upgrade transformers scipy
Transformers · model card · 🤗 Transformers Usage
from transformers import AutoProcessor, MusicgenForConditionalGeneration

processor = AutoProcessor.from_pretrained("facebook/musicgen-small")
model = MusicgenForConditionalGeneration.from_pretrained("facebook/musicgen-small")

inputs = processor(
    text=["80s pop track with bassy drums and synth", "90s rock song with loud guitars and heavy drums"],
    padding=True,
    return_tensors="pt",
)

audio_values = model.generate(**inputs, max_new_tokens=256)

Evidence level: Generic estimate only; no execution reproduced by this site.

Before downloading
  1. Check the exact checkpoint and license linked here.
  2. Confirm a compatible runtime and precision or quantization for your platform.
  3. Allow for context, batch size and any extra media components beyond the weights.

Technical details

What the source and the catalog store for this checkpoint. Missing fields are left out, never zero.

Checkpoint / version
facebook/musicgen-small
Task and modality
text-to-music · audio
Count and basis
0.3B · manufacturer model card
Released
audiocraft CHANGELOG.md (Meta FAIR) · read 2026-10-13source
Published repository size
5.8 GBNot runtime memory or a minimum download.files
Published components
300M autoregressive transformer · EnCodec 32 kHz tokenizer
Runtime and precision
No local execution verified.
Commercial use
Not allowed CC BY-NC 4.0 text (SPDX), the license the MusicGen card names for its weights · read 2026-10-18license text

The model card says weights are CC-BY-NC 4.0 while code is MIT; they are separate terms.

Sources

Identity and published specifications come from the checkpoint page. Benchmark scores belong to the named release and do not measure speed on your computer.

Catalog reviewed: 2026-09-22. Benchmark reviewed: 2026-10-20.