Skip to main content
POST

Overview

Takes a recording and makes another voice say it: the words, pacing and intonation stay, the voice changes. It is Speech to speech in the dashboard editor, with the same settings and price. Record yourself reading a script, then re-voice it with a library voice or one of your own clones.
  • Send a public https link to the recording in audioUrl: MP3, WAV, M4A, AAC, OGG, FLAC or WEBM, up to 5 minutes and 10 MB. It is downloaded in the request; local, private-network and plain http addresses are refused with 400.
  • voiceId is any voice from List Voices.
  • audio takes the same settings as Generate Speech: speed, stability, similarityBoost.
The MP3 is saved to your team’s voiceovers (counted against your storage) and url links to it for 10 minutes: download it right away.
Price: 5 credits per started minute of the recording (42 seconds: 5 credits; 70 seconds: 10). They are charged before generating and given back if it fails. BYOK teams use their own ElevenLabs key and are not charged.

Errors

Send an Idempotency-Key header to retry safely after a timeout.

Authorizations

x-api-key
string
header
required

Headers

Idempotency-Key
string

Makes the request safe to retry. 1-255 printable ASCII characters, one per operation (your job id, or a UUID you store), reused on every retry. Within 24 hours the same key with the same body answers with the first response and the header Idempotent-Replayed: true, without running or charging again. Keys are scoped to your team and the endpoint. 5xx answers and refusals before anything ran (401, 402, 403, 409, 429) are not kept. See Idempotency.

Required string length: 1 - 255
Pattern: ^[\x20-\x7E]+$

Body

application/json
voiceId
string
required

A voice from List Voices: its id (a library voice, or one of your team's own voices). Another team's voice, an unknown id, or one of your clones that is unavailable in your current key mode answers 400.

Maximum string length: 100
audioUrl
string<uri>
required

Public https link to the recording file itself: MP3, WAV, M4A, AAC, OGG, FLAC or WEBM, up to 5 minutes and 10 MB.

Maximum string length: 2000
audio
object

Voice settings, as in the editor. Each one is optional; a missing one takes its default. An unknown key answers 400.

Response

Speech generated

success
boolean
message
string
data
object