Audio Dialogue

Generate multi-speaker dialogue audio from a script.

POST/api/v1/audio/dialogue

Request Body

Parameter	Type	Required	Description
model	string	Yes	Model key (see models below)
params	object	Yes	Model-specific parameters (speakers, script, etc.)

Response

{
  "inferenceJobId": "ij_..."
}

Example

POST/api/v1/audio/dialogue

{
  "model": "elevenlabs-dialogue",
  "params": {
    "speakers": [
      { "label": "ALICE", "voiceId": "voice_alice123" },
      { "label": "BOB", "voiceId": "voice_bob456" }
    ],
    "script": "ALICE: Hey Bob, have you tried the new editor?\nBOB: Yes! The AI features are incredible.\nALICE: I know, right? It saves so much time.",
    "stability": 0.5
  }
}

Models

elevenlabs-dialogue

Multi-speaker dialogue generation via ElevenLabs. No prompt needed.

Parameter	Type	Required	Description
speakers	array	Yes	Array of { label: string, voiceId: string }. Each label must match a speaker in the script.
script	string	Yes	Dialogue script. Format: SPEAKER_LABEL: text (one line per speaker turn).
stability	number	No	Voice stability (0 to 1)

Attachments: None.

Script Format

ALICE: Hey Bob, have you tried the new editor?
BOB: Yes! The AI features are incredible.
ALICE: I know, right? It saves so much time.

Each line must begin with a speaker label that matches one of the entries in the speakers array, followed by a colon and the text to speak.