Audio Isolator separates vocals from background music, noise, and instrumentation in any audio file using AI. You give it a URL to an audio file and it returns a clean track with just the isolated voice.
It works on both speech and music. For podcast producers and interviewers, it cleans up recordings polluted by background noise. For musicians, it extracts a cappella vocals from a full song. The output is automatically stored as a downloadable file.
What you can do
- Remove background music and instrumentation from any audio file to isolate vocals
- Clean up podcast and interview recordings with background noise
- Extract a cappella vocals from songs for remixing or sampling
- Process any common audio format — MP3, WAV, M4A, and more
Who it's for
Podcast editors and audio producers cleaning up noisy recordings. Musicians extracting vocals for remixes or a cappella versions. Journalists and researchers separating speech from ambient sound in field recordings.
How to use it
- Call isolate_vocals with a publicly accessible URL to your audio file
- The tool processes the file and returns an audio_url with the clean isolated track
- For best results, use the highest-quality source file available — higher bitrate input produces better separation
Getting started
Pass any public audio URL to isolate_vocals — no configuration required. For noisy source material, run audio-transcriber afterward to get a clean transcript of the isolated vocals.
