Skip to main content
POST

Request

Headers

string
required
Bearer token with your API key
string
required
application/json

Body

string
required
The text to convert to speech (max 5000 characters)
string
ID of the voice to use
boolean
default:"true"
Enable audio enhancement
number
default:"1"
Speech speed multiplier (0.5 to 2.0)
number
Words per minute (optional). Target speaking rate in WPM. Range 60300. When omitted, the voice’s default WPM is used (returned as wpm by GET /voices); voices without a measured default fall back to 137 WPM.
number
default:"50"
Voice stability 0-100 (lower = more expressive, higher = more consistent)
number
default:"75"
Voice similarity 0-100 (how closely the output matches the source voice)

Response

The response is streamed as newline-delimited JSON (NDJSON). Each line contains a JSON object:

Chunk Object

string
Type of message: “chunk”, “complete”, or “error”
object
Contains the audio chunk data
string
Base64-encoded audio chunk
string
Error message (only for error type)

Use Cases

Streaming is ideal for:
  • Real-time applications: Voice assistants, chatbots
  • Long-form content: Articles, books, documents
  • Low latency: Start playing audio before generation completes
  • Progressive enhancement: Display text while generating audio