Separate Stems
AI-generatedSummary
Submits an audio file or URL to StemSplit to begin AI stem separation and returns the job details immediately without waiting for completion.
Inputs
- Input Source — Choose whether to upload an audio file from binary workflow data or supply a public URL.
- Input Binary Property (required) — Property name containing the binary audio data when Input Source is set to Binary File (defaults to 'data').
- Audio URL (required) — Publicly accessible web URL of the audio file to separate when Input Source is set to URL.
- Output Type — Specify which audio tracks to extract (Vocals + Instrumental, Vocals Only, Instrumental Only, Four Stems, or Six Stems).
- Quality — Select processing quality: Best (slowest, required for Six Stems), Balanced, or Fast.
- Output Format — Target file format for separated stems (MP3, WAV, or FLAC).
- Additional Options — Optional settings to override the file name or attach custom JSON metadata to the job.
Output shape
a single record
Returns initial job details immediately, including job ID, status, required credits, estimated seconds, and submitted options. Use the Get Job operation to retrieve separated stem URLs once processing finishes.
Examples
Example 1: Submit a binary MP3 for vocals and instrumental separation
Input Source: Binary File, Input Binary Property: data, Output Type: Vocals + Instrumental, Quality: Best, Output Format: MP3
Example 2: Submit an audio URL for 4-stem separation in WAV format
Input Source: URL, Audio URL: https://example.com/track.mp3, Output Type: Four Stems, Quality: Balanced, Output Format: WAV