Subtitles and transcription API
Send a file, get back a transcript, a subtitle file or a video with captions burned in. No GPU, no FFmpeg, no queue to run — one header and two requests.
How it works
- 1Create an API key in your profile.
- 2POST the file with the operation and its settings in one multipart request.
- 3Poll the job or receive a webhook, then download the result.
What you get
Transcripts and subtitle files
TXT, SRT, VTT, ASS and JSON with word timestamps — several formats from one request.
Captions on video
Burn styled captions into the picture or mux a soft track, in the same job as compression, resizing or trimming.
Broadcast-style controls
Characters per line, lines per cue, minimum and maximum cue length and reading speed, plus a custom vocabulary for names and jargon.
Built for automation
Async jobs with webhooks, an OpenAPI schema, a Postman-ready quickstart and an MCP server so Claude, Cursor or ChatGPT can call it.
Same thing over the API
One header, one multipart request with the file, then poll the job or take a webhook. The same operation you just tried in the browser.
Questions
Do I pay for idle time?
No. You pay per processed file, metered by processing time with a hard cap per file — the current numbers are on the pricing page.
Which inputs are accepted?
Video (MP4, MOV, MKV, WebM) and audio (MP3, WAV, M4A, OGG, FLAC), uploaded directly or by URL; large files go through presigned uploads.
Can I bring my own subtitles?
Yes — pass SRT, VTT or ASS by URL or inline and they are styled and burned in without recognition.
Is there a free way to test?
Every operation can be tried in the browser studio first, and the docs page has a key box that fills every example with your own key.