Privacy

When you fetch a transcript, the service writes one line to its request log: the video’s ID, the caption language, the time, whether it worked and how long it took. The line holds no IP address and no caption text, and it is deleted after 30 days.

The captions themselves are cached for 3 days, so the next request for the same video is quick.

Each transcript fetched also becomes a page in the public library, as below.

Your IP address is never written to our database or to the service’s logs. It is used only to count your requests against the rate limit, and each count is deleted after 15 minutes.

Like any web server, the one in front of this site keeps standard access logs, which include IP addresses.

There are no accounts and no cookies, and the site runs no analytics or advertising scripts.

Every transcript fetched here is added to the public library: a page for the video with its title, its channel, its thumbnail and the transcript. A library page never shows who fetched the video. The videos are public YouTube videos, and the transcripts come from the captions YouTube publishes for them.

The thumbnails load straight from YouTube’s image server, i.ytimg.com. Your browser asks for them without saying which page it is on, but like any server, YouTube’s sees your IP address.

Most pages also carry a summary, key points and answers to common questions, written by an AI model from the transcript and labelled as AI-generated. To write them, the transcript and the video’s title, channel and description are sent to DeepSeek, an AI provider. Nothing about you is sent.

Pages whose title, description or captions match our lists of adult or illegal terms stay out of the listings and out of search engines, and the worst matches are never published at all.

Every library page has a Request removal link, which opens an email to us with the video’s address in the subject. A removed video’s page is erased and answers “gone”, and the video is never added to the library again, even when someone fetches it later.