Create a custom voice persona from an audio file containing clear vocals. This is an async endpoint — submit the audio URL, receive a task_id, then poll for the result.

Credits cost: 5

Workflow:

1. Submit a publicly accessible audio URL (MP3 or WAV, >10s, single clear speaker)
2. Receive a task_id in the response
3. Poll GET /api/v1/sonic/task/{task_id} until the task completes
4. The completed task returns a persona_id
5. Use the persona_id with POST /api/v1/sonic/create (task_type: persona_music) to generate music with that voice

---

Polling Result

Use GET /api/v1/sonic/task/{task_id} to poll the result.

{
  "code": 200,
  "data": {
    "persona_id": "xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx",
    "name": "Voice Name"
  },
  "message": "success"
}