startTranscription()
startTranscription()
startTranscription({ language, model })
You can start the service by calling startTranscription() on your call object.
You can optionally include the following configuration options to control the transcription process:
| Name | Type | Description |
|---|---|---|
language | string | See Deepgram's documentation for language |
model | string | See Deepgram's documentation for model |
tier | string | This field is deprecated, use model instead |
profanity_filter | boolean | See Deepgram's documentation for profanity_filter |
punctuate | boolean | See Deepgram's documentation for punctuate |
endpointing | number or boolean | See Deepgram's documentation for endpointing |
redact | boolean or array | See Deepgram's documentation for redact |
extra | object | Specify additional parameters. See Deepgram's documentation for available streaming options |
Transcription message data is passed from the transcription service to the Daily call object via the "transcription-message" event. See the "transcription-message" event documentation for more info.
An additional configuration option is available which doesn't affect the transcription process itself, but gives you more detailed output data in the transcription messages:
| Name | Type | Description |
|---|---|---|
includeRawResponse | boolean | Whether Deepgram's raw response should be included in all transcription messages |