Demucs is Meta’s open-source music source separation project for splitting songs into stems such as vocals, drums, bass, and accompaniment. It fits agent workflows that need repeatable audio preprocessing before transcription, remixing, analysis, captioning, or archive preparation.
name demucs-music-source-separation-for-vocal-and-stem-extraction title Demucs Music Source Separation for Vocal and Stem Extraction description Demucs is Meta’s open-source music source separation project for splitting songs into stems such as vocals, drums, bass, and accompaniment. It fits agent workflows that need repeatable audio preprocessing before transcription, remixing, analysis, captioning, or archive preparation. category Media & Transcription framework Multi-Framework verification security_reviewed source https://github.com/facebookresearch/demucs tool_ecosystem {"github_repo":"facebookresearch/demucs","github_stars":9949} Demucs Music Source Separation for Vocal and Stem Extraction Demucs is Meta’s open-source music source separation project for splitting songs into stems such as vocals, drums, bass, and accompaniment. It fits agent workflows that need repeatable audio preprocessing before transcription, remixing, analysis, captioning, or archive preparation. Installation You can install this skill using any of these methods: OpenClaw skill installer ClawHub CLI Git clone into your skills directory Download and extract the skill folder manually Copy the skill folder from a local checkout Source Agent Skill Exchange