Transcript result schema

Work with transcript text, timestamps, duration, status, and job-level errors.

Job schema

Field	Type	Description
`id`required	uuid	Stable job identifier.
`status`required	string	processing, completed, or failed.
`filename`required	string	Original upload name or derived link label.
`created_at`required	date-time	UTC creation timestamp.
`duration_seconds`	number \| null	Detected media duration.
`text`	string \| null	Full transcript on completed get-one responses.
`segments`	array \| null	Timestamped transcript segments when available.
`error`	string \| null	Processing failure message when status is failed; map or sanitize it before end-user display.

{
  "data": {
    "created_at": "2026-09-03T10:12:00.000Z",
    "duration_seconds": 1842,
    "error": null,
    "filename": "interview.mp3",
    "id": "95c3fdd8-…",
    "segments": [
      { "start": 0, "end": 4.2, "text": "Welcome.", "speaker": 0 }
    ],
    "status": "completed",
    "text": "Welcome."
  }
}

Segments

Field	Type	Description
`start`required	number	Start offset in seconds.
`end`required	number	End offset in seconds.
`text`required	string	Text spoken during the segment.
`speaker`	number \| string	Optional speaker identifier when diarization is enabled; do not assume it is a display-ready label.

Field availability

Response	text	segments	error
List jobs	`null`	`null`	Only on failed summaries
Get processing job	`null`	`null`	`null`
Get completed job	string	array when available	`null`
Get failed job	`null`	`null`	string

JSON is the v1 output

The web product can create document and subtitle exports, but the public REST API currently returns JSON text and segments rather than export files.