Skip to main content
Takes a video URL, transcribes the speech with word-level timestamps, and returns the same video with styled subtitles burned in. Two presets: movie (clean cinematic lower-thirds) and tiktok (bold word-by-word social style). Works on any video you have a URL for — including videos generated here. (If you’re generating the video in the same breath, prefer make_video with captions: true — it chains both steps in one run.)

Arguments

Example

Response

202 with run_id; poll GET /api/v1/runs/{run_id}. On completed:
  • output[0] — the captioned MP4, on durable storage.
  • transcript — the full transcription text, stored on the run and full-text searchable via GET /api/v1/generations?q=....
On failed, the flat caption charge is refunded instantly.

Pricing

Captioning is a flat 15 credits per video, regardless of length or style — see Credits & refunds.