Sign in
Real-time Translation

Create a real-time translation session

Creates a new real-time translation session and returns WebSocket connection credentials. Connect to the returned url with a LiveKit client SDK using the token to start streaming audio and receiving translated audio + transcripts.

post/api/beta1/session

Request body

source_languagestring required

Source language code (e.g. "en-US"). 50+ languages supported.

target_languagestring required

Target language code (e.g. "es-ES"). 50+ languages supported.

voice_type'clone' | 'female' | 'male'

Voice type for translated audio. "clone" uses the speaker's voice.

Example request

{
  "source_language": "en-US",
  "target_language": "es-ES"
}

Response

Session created successfully

urlstring uri required

WebSocket URL for LiveKit connection

tokenstring required

JWT token for authenticating to the room

room_namestring required

Unique room identifier (format: api-<id>)

Changes