Video Transcript & Subtitle Extractor
Pricing
from $3.50 / 1,000 transcript extracteds
Extract transcripts, subtitles and captions from video URLs on 1,000+ platforms, including YouTube, Vimeo, TikTok, Instagram, X, Facebook, Twitch and TED. Native captions first, Whisper AI when none exist; output JSON, SRT, VTT, text or Markdown.
v0.11 (2026-11-12) — Fast YouTube caption route: one provider request per video (about 1-2 s instead of 30-60 s), parallel prefetch for playlists, existing extraction as automatic fallback. New option includeFullMetadata for views, likes, upload date and description. Speech fallback is capped at 10 minutes on the Free plan.
v0.5 (2026-10-16) — YouTube playlist and channel URLs expand into individual videos (maxVideosPerPlaylist, default 20); per-video time limit (perVideoTimeoutSecs, default 900s) skips a stuck video instead of timing out the run; blocked/transient probe errors retry once on a fresh residential session and then directly.
v0.4 (2026-08-05) — Make native-caption extraction the default so prefilled runs do not require an OpenAI credential; Whisper remains an explicit opt-in.
v0.3.7 (2026-07-29) — Apply maxDurationMinutes only to Whisper audio transcription, not native/manual/auto caption downloads. Long captioned lectures and streams now succeed without weakening the Whisper cost cap. Added regression tests for both paths.
v0.3 — Retries throttled subtitle downloads, falls through the ordered preferred-language list, emits only validated transcript rows with coupled PPE charging, and persists honest COMPLETE/PARTIAL/CONFIG_ERROR/UPSTREAM_FAILED outcomes. Failed URL diagnostics are stored in FAILURES rather than the result dataset.
