Configure Transcription
AI-generatedSummary
Configure the speech-to-text transcription settings for an ongoing sipgate AI Flow session by specifying a session ID and optionally changing the transcription provider, supported languages, and custom vocabulary.
Inputs
- Session ID (required) — The unique identifier of the AI Flow session to configure transcription for.
- Provider — Optionally switch to a different speech-to-text provider (Azure, Deepgram, ElevenLabs), or keep the current provider by selecting 'Keep Current'.
- Languages — Optionally specify a comma-separated list of 1 to 4 BCP-47 language codes (e.g., "de-DE,en-US") to update the transcription languages; leave empty to retain current languages.
- Custom Vocabulary — Optionally provide comma- or newline-separated phrases (up to 100 entries, each max 200 characters) to bias the transcription recognizer; leave empty to keep the current vocabulary.
Output shape
no body — success/failure only
Produces an action object summarizing the transcription configuration change to be sent to the AI Flow session. Throws an error if the session ID is missing or if input constraints are violated (e.g., too many languages or vocabulary entries). No data is directly returned by the node; it acts as an API action request.
Examples
Example 1: Change transcription provider to Azure with two languages and a custom vocabulary
Session ID is set to the current session's ID, provider changed to Azure, languages set to 'de-DE,en-US', and custom vocabulary includes 'sipgate, AI Flow, Düsseldorf'.
Example 2: Keep current provider, update languages to English US only, no vocabulary change
Session ID specified, provider set to 'Keep Current', languages updated to 'en-US', and custom vocabulary left empty.