How It Works

Complete transparency. No magic, no AI. Just clean engineering extracting what YouTube already has.

Step 01

Paste YouTube URL

Copy any YouTube video link and paste it into our input field. We support all standard YouTube URL formats including youtube.com and youtu.be links.

Input: https://youtube.com/watch?v=...

Step 02

Our system parses the URL to extract the unique 11-character video identifier. This is the key to fetching the right transcript.

Video ID: dQw4w9WgXcQ

Step 03

We use the youtube-transcript-plus library to query YouTube's internal timedtext API. This retrieves existing captions (manual or auto-generated).

API: youtubei/v1/player → captionTracks

Step 04

The transcript is parsed, formatted, and returned to you instantly. We don't store anything - the data flows directly from YouTube to your browser.

Response: { segments, fullText, count }

Transparency means being clear about our limitations

We don't use speech-to-text AI. We only extract existing captions YouTube already has. If a video has no captions, we can't help.

We don't download or process video files. We only work with text metadata - specifically the caption tracks.

Transcripts are fetched in real-time and never stored on our servers. Each request is completely independent.

We can only access public videos with captions enabled. Private, unlisted, or age-restricted videos won't work.

Modern, open-source technologies

React framework for production

Type-safe JavaScript

Utility-first styling

Beautiful components

Smooth animations

Transcript extraction