Voices

GET /v1/voices lists every voice you can use: the preset catalog plus your own clones. Works with an API key or a client token (?token=).

{
  "data": [
    {
      "id": "be7a1f2c-…",
      "slug": "leo",
      "name": "Leo",
      "description": "High-energy, motivational",
      "gender": "male",
      "accent": "American",
      "tags": ["energetic", "bold"],
      "type": "preset",
      "sampleUrl": "https://vaux.ng/samples/leo/neutral.mp3",
      "voiceId": "adff5dcb-…"
    }
  ]
}

Use slug, name or id with the SDK and /v1/tts. Use voiceId in raw WebSocket session.update messages.

Voice clones

POST /v1/voices/clone creates a private clone from 10–30 seconds of clean speech. Cloning requires verified consent from the voice owner; see the Voice Consent Policy. Cloning is rolling out; until it is enabled for your account the endpoint returns 403 feature_unavailable.