# facebookresearch/sam-audio

The repository provides code for running inference with the Meta Segment Anything Audio Model (SAM-Audio), links for downloading the trained model checkpoints, and example notebooks that show how to use the model.

Repository: https://github.com/facebookresearch/sam-audio
Canonical: https://ross.abutalabs.com/products/sam-audio
Language: Python
License: NOASSERTION
License Family: other
Last push: 2026-05-26T05:38:19+00:00

## Health v2 (maintenance only)
Score: 55/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 84, release rhythm 35, longevity 25
- inputs: {"age_days": 363, "days_push": 99, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3612, forks 330 (observed 2026-08-28T04:08:11.670405+00:00)

## What it is
SAM-Audio is Meta's foundation model for isolating any sound in audio using text, visual, or temporal prompts. This repository provides inference code, model checkpoints via Hugging Face, and example notebooks for separating specific sounds from complex audio mixtures.

## Use cases
- separate a specific sound from an audio mixture using a text description
- isolate a sound in audio based on visual cues from video
- extract audio segments within a given time span
- run inference with the SAM Audio foundation model
- split an audio file into a target sound and residual audio
- download and experiment with Meta's audio segmentation checkpoints

## When to choose
- you need prompt-driven audio source separation with natural language descriptions
- you want state-of-the-art sound isolation from a research-grade foundation model
- you have a CUDA GPU and want to integrate audio separation into a Python pipeline

## When to avoid
- you need a lightweight CPU-only audio separation tool
- you want a ready-made end-user application rather than a Python library
- you cannot obtain access to the gated Hugging Face checkpoints

## Facets
- artifact type: library
- maturity: active
- function: audio-processing, machine-learning, deep-learning, llm-inference
- domain: machine-learning, artificial-intelligence, speech-processing
- platform: python, windows
- tags: audio-separation, source-separation, foundation-model, segment-anything, meta-ai, text-prompting, audio-visual, inference, audio, gpu, linux, macos

## Member repositories
- facebookresearch/sam-audio (main) score 55

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:11.670405+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 2026-08-29T18:33:38.386186+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/sam-audio (fetched 2026-08-28T04:08:11.670405+00:00, sha 359abc70c1bb)
- Data as of 2026-08-30T08:39:29.467469+00:00.