Voice Clone Music APICustom Voice Personas
Turn a reference voice into a reusable persona, then generate songs sung in that voice. Create a persona from an audio URL, sing your lyrics with it, poll for the result, and download — with API keys, free trial credits, pricing, and production-ready docs.
What is the Voice Clone Music API?
It's a two-step flow: first create a voice persona from a reference audio URL, then generate songs in that voice with task_type: persona_music. The same persona can be reused across many generations for a consistent artist sound.
- Clone from Audio
Create a persona from a public MP3/WAV
audio_url— no model training to manage.- Consistent Artist
Reuse one persona across generations so an app's songs share a recognizable voice.
- Durable Option
For long-lived personas, use the upload + persona two-step so they don't orphan over time.
API Examples
Create a persona from reference audio, then generate a song in that voice.
1curl -X POST 'https://api.musicapi.ai/api/v1/sonic/create-voice' \
2--header 'Authorization: Bearer YOUR_API_KEY' \
3--header 'Content-Type: application/json' \
4--data '{
5 "audio_url": "https://example.com/reference-voice.mp3"
6}'Personas from /create-voice can expire; for durable personas use /sonic/upload then /sonic/persona. Ensure you have the rights to any reference voice you submit.
Why Build Voice Personas on MusicAPI?
Reusable Voice
One persona, many songs in a consistent voice.
Async + Fast
Generation returns a task id; songs finish in 1-3 min.
Auto-Refunds
Failed upstream tasks are refunded; polling is free.
Pairs with Lyrics
Generate lyrics, then sing them in your persona.
Voice Clone Music API: Pricing, Free Access & API Keys
Developers adding voice personas usually want three answers: how to test for free, how to get an API key, and how cost scales after the prototype works.
Free trial
Use free credits to test persona creation and generation before choosing a paid plan.
Pricing
Credit-based. Estimate per-song cost on the pricing page.
API keys
Generate a bearer token in your account, keep it server-side, and call the create-voice and create endpoints.
Ready to Build with Custom Voices?
Add voice personas to your app with a couple of endpoints. Start free, then scale with predictable credit pricing.
Frequently Asked Questions
Everything you need to know about voice personas with the API.
What is a voice clone music API?
A voice clone music API turns a reference voice into a reusable persona, then generates songs sung in that voice. With MusicAPI you create a persona from a reference audio URL, then call the generation endpoint with task_type 'persona_music' to sing your lyrics in that voice.
How do I clone a voice with the API?
POST a publicly accessible MP3/WAV audio_url to https://api.musicapi.ai/api/v1/sonic/create-voice. It returns a persona you can reuse. Then POST to /api/v1/sonic/create with task_type 'persona_music', the persona_id, your lyrics in prompt, and a model version (mv); poll GET /api/v1/sonic/task/{id} for the song.
What audio do I need to create a persona?
A publicly accessible MP3 or WAV URL of the reference voice. Make sure you have the rights to use the voice you submit.
Are personas permanent?
Personas from /create-voice are bound to backend resources and can become unavailable over time. For durable personas, use the two-step flow: upload your audio via /api/v1/sonic/upload, then train via /api/v1/sonic/persona using the returned clip_id — those are trained on shared resources and don't orphan the same way.
How do I generate a song in the cloned voice?
Call /api/v1/sonic/create with task_type 'persona_music', the persona_id, and your lyrics in the prompt field plus a model version. The result is an async task — poll the task endpoint until the song is ready.
How is voice clone pricing handled?
Persona creation and song generation use credit-based pricing. Test with free trial credits, estimate cost on the pricing page, and upgrade as volume grows. Failed upstream tasks are auto-refunded and polling task status is free.
Do the results include commercial rights?
Audio generated through the API comes with commercial rights included. You are responsible for having the rights to any reference voice you submit.