# Audio-AGI/AudioSep

Official implementation of "Separate Anything You Describe"

Repository: https://github.com/Audio-AGI/AudioSep
Canonical: https://ross.abutalabs.com/products/audiosep
Homepage: https://audio-agi.github.io/Separate-Anything-You-Describe/
Language: Python
License: MIT
License Family: permissive
Last push: 2024-11-26T20:33:39+00:00

## Health v2 (maintenance only)
Score: 28/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 80
- inputs: {"age_days": 1121, "days_push": 645, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1929, forks 156 (observed 2026-08-28T04:05:55.911604+00:00)

## What it is
AudioSep is the official implementation of the 'Separate Anything You Describe' foundation model for open-domain, language-queried audio source separation. It separates a target sound from an audio mixture given a natural language query, supporting tasks like audio event separation, instrument separation, and speech enhancement.

## Use cases
- separate a specific sound from an audio mixture using a text description
- separate musical instruments from a mixed track
- enhance speech by removing background noise via a text query
- zero-shot separation of unseen audio events
- run language-queried audio source separation in a Python pipeline

## When to choose
- you need text-query-driven audio source separation with zero-shot generalization
- you want a research-grade foundation model for audio event, instrument, or speech separation
- you need a Python API with pretrained checkpoints and Hugging Face/Colab integration

## When to avoid
- you need production-grade, low-latency real-time audio processing
- you lack GPU resources or don't want to manage model checkpoints
- you need simple classical DSP filtering rather than learned separation

## Facets
- artifact type: library
- maturity: active
- function: audio-processing, machine-learning, deep-learning, nlp
- domain: machine-learning, deep-learning, artificial-intelligence
- platform: python, cross-platform
- tags: audio-source-separation, text-queried-separation, foundation-model, speech-enhancement, zero-shot, lass, audio, gpu

## Member repositories
- Audio-AGI/AudioSep (main) score 28

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:55.911604+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 2026-08-30T03:08:51.520242+00:00, confidence not recorded.
  - readme: https://github.com/Audio-AGI/AudioSep (fetched 2026-08-28T04:05:55.911604+00:00, sha 7f1a9e06e848)
  - homepage: https://audio-agi.github.io/Separate-Anything-You-Describe/ (fetched 2026-08-29T10:48:20.467901+00:00, sha aa5fc2c0d7c1)
- Data as of 2026-08-30T08:39:29.467469+00:00.