Video is where most knowledge lives on the internet. Tutorials, interviews, research, commentary — all of it is stuck in a format agents can't read. Stophy fixes that.
One API call gets you transcripts with timestamps, threaded comments, channel metadata, search results, and playlist contents — all as structured JSON designed for tool calls.
1curl -X POST https://api.stophy.dev/v1/video \2 -H "Authorization: Bearer $STOPHY_API_KEY" \3 -H "Content-Type: application/json" \4 -d '{"videoUrl": "https://youtube.com/watch?v=...", "type": "transcript"}'We built this because our own agents kept running into the same wall. Getting video data meant either hitting YouTube's official API (quota-limited, doesn't give you transcript text) or building scrapers that break regularly. We solved it for ourselves and decided to ship it.
What the API covers
- Transcripts — full text with timestamps, broken into segments. Auto-detection of whether captions are human-written or auto-generated.
- Comments — paginated threads with likes, replies, author metadata, and whether the creator pinned or hearted each one.
- Search — keyword search with filters for content type, duration, upload date, and sort order.
- Channels — profile data, video catalog, Shorts, playlists, and community posts.
- Playlists — full item lists with pagination.
Credits
Every API call costs one credit, regardless of endpoint or response size. New accounts get 100 credits on signup. No subscriptions, no monthly resets, credits don't expire.
Get started
Sign up at stophy.dev, grab your key from the dashboard, and make your first request. The docs have working examples for every endpoint in JavaScript, Python, and curl.