Skip to main content

Alternative API modes

You can enable captions only or captions with translation modes by adjusting the task options.
Depending on your input_stream, use WebSockets or WebRTC DataChannel to receive caption messages.
Refer to the recommended settings for exact option values.

Captions only​

All input_stream and transcription options are supported.

{
  "input_stream": {/*...*/},
  "output_stream": null, // set to `null`
  "pipeline": {
    "transcription": {/*...*/},
    "translations": [], // set to an empty list
    "allowed_message_types": [
      // you will only receive messages of these types
      "partial_transcription",
      "validated_transcription"
    ]
  }
}

Captions with translations​

To enable this mode, set output_stream to null. You may omit the speech_generation options if not needed.
All standard task options are supported — see the recommended settings.

{
  "input_stream": {/*...*/},
  "output_stream": null, // set to `null`
  "pipeline": {/*...*/}
}