YouTube Transcript Extractor

How it works: YouTube doesn't expose a public CORS-friendly transcript API, so requests are routed through this site's own Cloudflare Pages Functions. Video metadata is fetched via YouTube's oembed endpoint and the list of caption tracks via /api/timedtext?type=list — neither sits behind the "confirm you're not a bot" wall that gates the InnerTube player API. Caption XML is then pulled through a small whitelist proxy. Nothing is stored. Only user-uploaded captions are exposed; videos with auto-generated (ASR) captions only can't be extracted this way.