Speech-to-text API with word-level timings and SRT

Send an audio or video file (or a public link to one) and receive a transcript with word-level timings as JSON and as SRT subtitles, optionally tagging laughter, music and other sounds.

Who it is for

Apps that caption or search video and audio
Meeting and interview tools
Editors who need subtitle files

How it works

Upload the file or pass file_url (mp3, wav, m4a, webm, flac and more, up to 200 MB).
Poll the task or use a webhook.
Download output.json_url for word timings or output.srt_url for subtitles.

Endpoints

Method	Path	What it does
`POST`	`/v1/audio/transcriptions`	Start a transcription task
`GET`	`/v1/tasks/{id}`	Get json_url and srt_url

Models

Pass one of these as model (or voice_id for voices). The full, current list with prices is at /v1/models.

ai33/elevenlabs-scribe

Example

curl https://omni-api.rcinc.app/v1/audio/transcriptions \
  -H "Authorization: Bearer $OMNI_KEY" -H "content-type: application/json" \
  -d '{"file_url":"https://example.com/interview.mp3","tag_audio_events":true}'

Pricing

Charged per job from the provider's actual cost; the task shows the charge. See the pricing page for the full comparison.

Limits

Direct uploads through the API are limited to about 4 MB; use file_url for larger files.
Result links expire, so download them promptly.