What it does
You can paste captions when YouTube’s undocumented caption endpoints fail, a particular risk on hosted deployments.
Apps & data pipelines
Transcript Lens
Navigate a YouTube transcript by topic, chapter, and semantic heat map without asking Jev to rewrite it.
Bookmark: Transcript Lens
Keep this in your collection.
Leave a noteWhat would you try with this?
: Transcript Lens
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
How you can use it
Your developer can look at the Transcript Lens code to build a video reading tool. The project shows how to break a long script into smaller topics. You can paste a video script into the app. It then sends sections of the text to an outside AI service called Jev. Jev labels important parts and changing topics without rewriting the original words.
This analysis needs an access key that connects the app to Jev. If you do not have this key, you can still paste and read text. The app tries to download captions directly from YouTube, but this connection often fails. Your developer can borrow the code that lets you paste the text yourself when that happens.