What is Suno Voice Creation?
Suno Voice Creation turns a recording of a real singer into a reusable voice for Suno music generation. You send two audio URLs: voice_audio_url, a sample of the singing voice, and verification_audio_url, the same person reading a short phrase. The phrase comes from GET /v1/resources/suno-ai/voice/verification-phrase, which returns a phrase_id and the phrase_text to read aloud. Add a name, and optionally a description, genre styles, and a singer_skill_level. When the task completes it returns a persona_id. Every later Suno music request that passes that ID as persona_id is sung in the custom voice. This is the endpoint to use when an artist or a brand wants the same voice across many tracks.
Key features of Suno Voice Creation
Four features shape how voice creation works.
Built-in ownership check
The verification recording and phrase_id confirm the voice belongs to the person submitting it. If the spoken phrase does not match, the task fails with a clear rejection message.
Reusable persona_id
The completed task returns a persona_id. Store it once and pass it to Suno music generation whenever a track should use that voice.
Trim the source recording
vocal_start_s and vocal_end_s mark where the vocal sits inside voice_audio_url, so you can upload a longer file and point at the part with clean singing.
Style and skill metadata
Add genre tags in styles (for example "funk edm"), a free-text description, and a singer_skill_level from Beginner to Professional to describe the voice.
Best for
Artist releases
Keep one recognizable voice across an album or a series of singles generated with Suno.
Brand jingles
Record a brand singer once and generate new jingles and sung intros in the same voice.
Creator tools
Let users record themselves in your app and then generate songs sung in their own voice.
Game and story characters
Give a recurring character a consistent singing voice across in-game songs and cut scenes.
Use cases
Build an onboarding flow in a music app: fetch a verification phrase, have the user record a short singing sample and the phrase in the browser, upload both to your storage, call suno-ai/voice, and save the returned persona_id on the user profile. Every song the user generates after that passes the stored ID to Suno music generation. For a label workflow, create one voice per artist and keep the IDs in your catalog database, then batch-generate demos in each voice. Pair it with Suno Add Vocals when you already have an instrumental and want the custom voice on top.
API examples
The full flow, including the verification-phrase step, is at docs.unifically.com/models/audio/suno/voice.
curl -X POST https://api.unifically.com/v1/tasks \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "suno-ai/voice",
"input": {
"voice_audio_url": "https://example.com/my-voice.wav",
"verification_audio_url": "https://example.com/my-verification.wav",
"phrase_id": "35989bb7-9fe5-4be2-b7d2-ef8124ae3e8b",
"name": "My Voice",
"description": "Warm male vocal",
"styles": "funk edm",
"singer_skill_level": "Intermediate"
}
}'
The response contains a task_id. Poll GET /v1/tasks/<task_id> or set a callback_url on the request to receive the persona_id once the voice is created.
FAQs
People also ask
A completed task with a persona_id, plus the voice name, an image_url, and the styles you set. Pass that persona_id as persona_id in a Suno music generation request and the track is sung in your custom voice.
To confirm the person in the voice sample is the person creating the voice. First call GET /v1/resources/suno-ai/voice/verification-phrase to get a phrase_id and phrase_text, record someone reading that text, and send the recording as verification_audio_url together with the phrase_id.
The task ends with status failed and a message such as "Voice verification rejected: didnt_say_verification_phrase". Re-record the phrase clearly, check that it matches the phrase_id you sent, and submit again.
Yes. Set vocal_start_s and vocal_end_s (in seconds) to mark the section of voice_audio_url that contains the vocal. vocal_start_s defaults to 0, and the end is detected automatically if you leave it out.
It tells Suno how polished the source singer is. Options are Beginner (the default), Intermediate, Advanced, and Professional. Set it to match the recording you upload.
