VoiceChange API

Change the voice of a recording, turn text into speech or make a song cover from your own code or from Zapier. Jobs use your account’s credits at the same prices as the site.

Zapier

In Zapier, add the VoiceChange app and connect it with an API key. Actions: Change Voice, Text to Speech and Make Song Cover. Trigger: Job Finished. Each action waits for its job and hands the audio file to the next step.

API keys

Make a key under Account → API keys. It is shown once, so keep it somewhere safe. Send it with every request:

Authorization: Bearer vc_your_key

Revoke a key on the same page and anything using it stops working at once.

Endpoints

Base URL: https://voicechange.io/api/v1. Requests and responses are JSON. Full reference as OpenAPI 3.1: voicechange.io/openapi.json.

GET /meThe key and the account’s credit balance.
GET /voicesVoices you can use. ?kind=cover lists only voices that can sing.
POST /jobsStart a job (below). Answers 201 with the job.
GET /jobs/{id}A job. Has audio_url once status is finished.
GET /jobsYour API jobs, newest first. ?status=finished, ?limit= (max 100).

Start a job

kindvoice (change a recording’s voice, default), tts (text to speech) or cover (song cover).
voice_idAn id from GET /voices.
audio_urlFor voice and cover: a public link to mp3, wav, m4a, ogg, flac or webm, up to 15 MB and 10 minutes. Or send the file as multipart field audio.
textFor tts: up to 1,000 characters.
pitchOptional, −24 to 24 semitones. auto_pitch (default true) matches the voice’s range.
callback_urlOptional https URL. When the job finishes or fails we POST the job JSON to it.
curl https://voicechange.io/api/v1/jobs \
  -H "Authorization: Bearer $VOICECHANGE_KEY" \
  -H "Content-Type: application/json" \
  -d '{"kind": "tts", "voice_id": "VOICE_ID", "text": "Hello there!"}'
{
  "id": "a260984b-…",
  "kind": "tts",
  "status": "finished",
  "voice_name": "Adam",
  "credits": 1,
  "audio_url": "https://…  (signed link, valid 1 hour)",
  "audio_format": "wav",
  "share_url": "https://voicechange.io/s/a260984b-…",
  "error": null,
  "created_at": "2026-11-17T20:00:24Z",
  "finished_at": "2026-11-17T20:00:36Z"
}

status goes queued → processing → finished or failed. Most jobs finish in under 30 seconds. Poll every few seconds or use callback_url.

Credits and limits

  • 1 credit per 10 seconds of audio, Premium voices 1 per 5 seconds. Text to speech counts about 14 characters per second.
  • Credits are held when a job starts and refunded if it fails. A failed job shows credits: 0.
  • Up to 30 new jobs a minute and 5 running at once per account.

Errors

Errors answer with an HTTP status and {"error": "code", "message": "…"}.

401 invalid_api_keyMissing, wrong or revoked key.
402 no_creditsNot enough credits. Get more.
400 / 404 / 415Something about the request: the message says what.
429Too many jobs. Wait a moment.

Use

The same terms of use apply as on the site: don’t use voices to deceive, defraud or harass anyone. Questions: [email protected].