POST/api/global-chatbot/demo/voice-transcribe

Converts a spoken audio recording from the demo page into text (speech-to-text), if voice is enabled. Demo counterpart to widget.voice_transcribe, bound to the Zentor marketing site instead of a tenant widget.

Auth & Security

No authentication required

Origin check fixed to https://zentor-app.de / https://www.zentor-app.de (not configurable, replacement for Widget-Token on the demo path).

Idempotent: No

Parameters

visitor_id(body, string, required)— Must be a valid UUID
audio(body, file, required)— multipart/form-data audio file

Example Response

{"ok":false,"error":{"code":"INVALID_VISITOR_ID","message":"visitor_id muss eine gültige UUID sein."}}

Error Codes

400 INVALID_VISITOR_ID — visitor_id is missing or not a valid UUID.
403 ORIGIN_DENIED — Request origin is not zentor-app.de.
413 AUDIO_TOO_LARGE — Audio file exceeds the limit.
415 AUDIO_TYPE_UNSUPPORTED — Audio format not supported.

Live Test Proof

Negative cases (400 INVALID_VISITOR_ID with correct origin header, 403 ORIGIN_DENIED) live-verified. The success case (real transcription) cannot be triggered outside the production domain, since the origin check is hardcoded to zentor-app.de — documented analogously to widget.voice_transcribe.

← Demo Chatbot and SSO