Media · Transcription · Start a video/audio transcription job (step 1 of 3)
Step 1 — submit a media URL for speech-to-text transcription.
The response is only {id}. It does not contain the transcript
and it does not contain any status flag. The STT pipeline runs
asynchronously and updates the row later via provider webhooks.
Next steps:
- Step 2: poll
getTranscriptionwith the returnediduntiljob_status="DONE"(or"FAILED"). - Step 3: optionally call
downloadTranscriptionFileto get VTT/SRT.
Authentication
Header authentication of the form Api-Key <token>
Bearer authentication of the form Bearer <token>, where token is your auth token.
Request
BCP-47 language code of the source audio. Defaults to "en-US".
Response
Transcription completed (the apikit middleware blocks until
terminal). Response is the full TranscriptionResponse when
job_status="DONE", or an {error, id, status, message} envelope
when the job fails or times out.
URL of the captioned/burned-in video, if produced.
Transcription payload. When job_status=COMPLETED, contains
language_code and phrases (each with speaker, text, and
words[{text, start, end, confidence}] in milliseconds).
Source metadata (e.g. {source, source_id}).
Lifecycle state of the transcription job.
Terminal: DONE (success) and FAILED. All other values are
non-terminal — keep polling.
Project status (PUBLIC, PRIVATE, ARCHIVED).
