Skip to main content
POST
Clone a voice

Authorizations

Authorization
string
header
required

Pass your API key as a Bearer token: Authorization: Bearer sk_996_....

Body

multipart/form-data
clip
file
required

Reference audio to clone from. WAV recommended; FLAC, OGG, MP3, and M4A also work. 1–300 s, under 20 MB. Only the first 20 s of speech (after silence trimming) is used.

name
string
required

A name for the cloned voice.

language
string

A well-formed BCP-47 language tag, e.g. en-US, en-GB, es-MX, pt-BR, nl-NL. Stored as metadata (used for listing and filtering); it doesn't change synthesis. The value is canonicalized on save (en-us → en-US, pt-br → pt-BR). Malformed tags (en_US, en-USA, english_, 123) are rejected with invalid_language (400).

Example:

"en-US"

description
string

A description of the voice.

transcript
string

The exact words spoken in the clip. With a transcript, the clone follows the speaker, including their pace, more closely. It must match the audio exactly, and it's ignored if the clip has 20 s or more of speech after silence trimming, so send a clip under 20 s when you use it. Not a Cartesia SDK field: pass it via extra_body (Python) or as an extra key with a type cast (TypeScript).

Response

The newly created voice.

id
string<uuid>
Example:

"5148a2db-5f7c-480c-8f7c-39c6414a4d5d"

is_owner
boolean

true if your org owns the voice (you cloned it).

Example:

false

is_public
boolean

true for shared library voices.

Example:

true

name
string
Example:

"Jessica"

description
string
Example:

"Polished, quick New York professional around thirty-five, warm underneath the hustle. Jessica makes a stressful thing feel manageable."

gender
enum<string>
Available options:
masculine,
feminine,
gender_neutral
Example:

"feminine"

language
string

BCP-47 dialect tag (e.g. en-US, nl-NL).

Example:

"en-US"

country
string

Country code.

Example:

"US"

created_at
string<date-time>

When the voice was created.

Example:

"2026-08-14T18:26:40.876Z"

preview_file_url
string

Only present when the request sends expand[]=preview_file_url. A signed, browser-playable URL of the voice's preview.