version: "1.0.0" name: separate-speakers description: Diarize a recording and split it into per-speaker audio tracks. license: MIT

Skill: separate-speakers

Identify who spoke when in a multi-speaker recording, and optionally split it into one clean audio track per speaker. AudioDiarize is AudioPod's speaker separation engine.

When to use

Invoke when the user wants to:

• Label a conversation by speaker ("Speaker 1", "Speaker 2", …)
• Extract a single participant's audio from an interview or meeting
• Prepare per-speaker tracks for editing or transcription

API

Diarization (labels + timeline):

POST https://api.audiopod.ai/api/v1/speaker/diarize

Per-speaker audio tracks:

POST https://api.audiopod.ai/api/v1/speaker/separate

Headers:

• X-API-Key: ap_your_api_key
• Content-Type: multipart/form-data

Body (form-data):

• file: audio or video file (or url to a public file)
• min_speakers / max_speakers: optional bounds (1–20)

Response

{ "job_id": 12345, "status": "PENDING" }

Poll GET /api/v1/speaker/jobs/{job_id} until COMPLETED for the per-speaker segments and download URLs.

Pricing

Start free. See https://www.audiopod.ai/developers/pricing for per-minute rates.