Skip to main content
PATCH
/
api
/
v1
/
bots
/
{object_id}
/
transcription_settings
Update Transcription Settings
curl --request PATCH \
  --url https://api.example.com/api/v1/bots/{object_id}/transcription_settings \
  --header 'Content-Type: application/json' \
  --data '
{
  "deepgram.language": "<string>",
  "deepgram.model": "<string>",
  "deepgram.keyterms": [
    {}
  ],
  "deepgram.redact": [
    {}
  ],
  "deepgram.use_eu_server": true,
  "openai.model": "<string>",
  "openai.language": "<string>",
  "openai.prompt": "<string>",
  "meeting_closed_captions.google_meet_language": "<string>",
  "meeting_closed_captions.zoom_language": "<string>",
  "meeting_closed_captions.teams_language": "<string>"
}
'
Updates the transcription settings for a bot. This endpoint is not explicitly defined in the current OpenAPI spec but follows the pattern of other bot settings endpoints.
This endpoint allows you to modify transcription settings mid-meeting for supported providers.

​
Authentication

Requires API key authentication via the Authorization header.

​
Path Parameters

​
object_id
string
required
Bot ID (e.g., bot_xxxxxxxxxxx)

​
Request Body

The transcription settings object supports various providers including Deepgram, Gladia, OpenAI, AssemblyAI, and more.

​
Deepgram Settings

​
deepgram.language
string
Language code for transcription. Defaults to ‘multi’ for automatic language detection. See Deepgram docs for available languages.
​
deepgram.model
string
Model to use for transcription. Defaults to ‘nova-3’.
​
deepgram.keyterms
array
Array of key terms to improve recall. Only available for nova-3 model in English.
​
deepgram.redact
array
Array of redaction types: pci, pii, or numbers.
​
deepgram.use_eu_server
boolean
Whether to use the EU server for transcription.

​
OpenAI Settings

​
openai.model
string
required
OpenAI model: gpt-4o-transcribe, gpt-4o-mini-transcribe, or gpt-4o-transcribe-diarize.
​
openai.language
string
ISO-639 language code for transcription.
​
openai.prompt
string
Optional prompt for the transcription.

​
Meeting Closed Captions

​
meeting_closed_captions.google_meet_language
string
Language code for Google Meet closed captions (e.g., ‘en-US’).
​
meeting_closed_captions.zoom_language
string
Language for Zoom closed captions (e.g., ‘English’, ‘Spanish’).
​
meeting_closed_captions.teams_language
string
Language code for Teams closed captions (e.g., ‘en-us’).

​
Response

Returns a success status when settings are updated.

​
Example Request

curl -X PATCH https://app.attendee.dev/api/v1/bots/bot_weIAju4OXNZkDTpZ/transcription_settings \
  -H "Authorization: Token YOUR_API_KEY_HERE" \
  -H "Content-Type: application/json" \
  -d '{
    "deepgram": {
      "language": "en",
      "model": "nova-3",
      "keyterms": ["product", "revenue", "KPI"]
    }
  }'

​
Example with Redaction

curl -X PATCH https://app.attendee.dev/api/v1/bots/bot_weIAju4OXNZkDTpZ/transcription_settings \
  -H "Authorization: Token YOUR_API_KEY_HERE" \
  -H "Content-Type: application/json" \
  -d '{
    "deepgram": {
      "language": "multi",
      "redact": ["pii", "pci"]
    }
  }'

Build docs developers (and LLMs) love

Get started for freeTalk to us