Clone a voice

POST/voices/clone

Clone a voice from an audio recording. Send audio as a base64-encoded string. The cloned voice will be available in GET /voices.

Request

curl -X POST "https://www.getrevox.com/api/voices/clone" \
  -H "Authorization: Bearer $REVOX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "audio": "UklGRiQAAABXQVZFZm10IBAAAAABAAEAQB8AAIA+AAACABAAZGF0YQAAAAA=",
    "contentType": "audio/wav",
    "name": "Marie",
    "language": "fr"
  }'

const res = await fetch("https://www.getrevox.com/api/voices/clone", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.REVOX_API_KEY}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    audio: "UklGRiQAAABXQVZFZm10IBAAAAABAAEAQB8AAIA+AAACABAAZGF0YQAAAAA=",
    contentType: "audio/wav",
    name: "Marie",
    language: "fr",
  }),
});
const data = await res.json();
console.log(data);

import os

import requests

response = requests.post(
    "https://www.getrevox.com/api/voices/clone",
    headers={"Authorization": f"Bearer {os.environ['REVOX_API_KEY']}"},
    json={
        "audio": "UklGRiQAAABXQVZFZm10IBAAAAABAAEAQB8AAIA+AAACABAAZGF0YQAAAAA=",
        "contentType": "audio/wav",
        "name": "Marie",
        "language": "fr",
    },
    timeout=30,
)
response.raise_for_status()
print(response.json())

audio is the recording in base64 (shortened here). Use the returned cartesiaVoiceId as voice.id, with provider cartesia.

Try itSend this request live

Response

Created

{
  "id": "6f9a3c1e-2b7d-4e58-9a10-3d2c4b5e6f70",
  "cartesiaVoiceId": "a0e99841-438c-4a64-b679-ae501e7d6091",
  "name": "Appointment reminder",
  "language": "en",
  "createdAt": "2026-11-05T09:00:00Z"
}

{
  "message": "Request validation failed",
  "bodyErrors": {
    "issues": [
      {
        "path": ["audio"],
        "message": "Invalid input: expected string, received undefined",
        "code": "invalid_type"
      }
    ]
  }
}

{ "error": "Unauthorized, invalid API key" }

{ "error": "An upstream provider failed" }

Request body

application/json

audiostringrequired
Base64-encoded audio data (max ~7.5MB decoded)
at most 10000000 characters
contentTypeenum (6 values)required
MIME type of the audio
Possible values"audio/webm""audio/wav""audio/mpeg""audio/ogg""audio/mp4""video/webm"
namestringrequired
Name for the cloned voice
at least 1 character
languageenum (8 values)required
Language of the voice
"en"
English
"fr"
French
"es"
Spanish
"de"
German
"it"
Italian
"pt"
Portuguese
"ru"
Russian
"zh"
Mandarin Chinese

Response

201

idstring
Revox's internal ID of the cloned voice record. The other endpoints use cartesiaVoiceId instead.
cartesiaVoiceIdstring
ID of the voice at Cartesia, which hosts cloned voices: the voice's id in GET /voices and the {id} of DELETE /voices/{id}. Give the voice to an assistant with voice.provider 'cartesia' and this value as voice.id.
namestring
Name of the cloned voice.
languagestring
Language of the voice.
createdAtstring
When the voice was cloned (ISO 8601).

Errors

400Invalid request: a field or parameter failed validation{ "message"?: string, "error"?: string, "bodyErrors"?: {…} }
401Missing or invalid API key{ "message"?: string, "error"?: string, "bodyErrors"?: {…} }
502An upstream provider failed{ "message"?: string, "error"?: string, "bodyErrors"?: {…} }

Error bodies carry an error (or message) string; a 400 adds bodyErrors.issues. See Errors.