YouTube Transcript Exporter

Export YouTube transcripts to LLMs or download them as text files. Easy customizable via settings panels. Additional features: persistent progress bar with chapter markers; display remaining time based on playback speed; auto-open chapter panels; links in the header; custom CSS; hide nav bar; color-coded borders on the home page based on video age.

目前為 2025-01-18 提交的版本,檢視 最新版本

作者
TimMacy
評價
0 0 0
版本
7.3.5
建立日期
2024-12-24
更新日期
2025-01-18
尺寸
133.1 KB
授權條款
GNU AFFERO GENERAL PUBLIC LICENSE-3.0
腳本執行於

YouTube Transcript Exporter

This script adds four buttons to the YouTube header to facilitate exporting a video's transcript to LLMs or downloading it as a text file. The first button copies the transcript to the clipboard before opening NotebookLM, ideal for quickly generating podcasts. Button two includes a prompt to summarize the video and opens ChatGPT. The third button downloads the transcript as a text file, while the last button opens the settings panel.

Note: Due to the Same-Origin Policy (SOP) in web development, JavaScript cannot directly interact across different websites: i.e. while the script can copy the transcript to the clipboard from YouTube and open a link (e.g. ChatGPT or NotebookLM), the copied transcript must be pasted manually into the text field via CTRL+V (Windows/Linux) or CMD+V (macOS).

Features

  • Core Script Functionality
    • Button One: Copies the transcript and opens NotebookLM.
    • Button Two: Copies the transcript with a prompt for summarizing before opening ChatGPT.
    • Button Three: Downloads the transcript as a text file.
    • Button Four: Opens the settings panel.
    • Transcript Formatting: Includes timestamps, chapter headers, video title, channel name, upload date, and URL.
  • Additional Features
    • Display Remaining Time: Shows the remaining playback time below the video, adjusted for playback speed.
    • Persistent Progress Bar: Keeps the progress bar with chapter markers visible below the video—even in fullscreen.
    • Auto Open Chapters: Automatically opens the chapter panel if available.
    • Links in Header: Up to seven links can be added next to the logo and the Left Navigation Bar can be hidden.
    • Customize CSS: Square Design, Compact Layout, Number of Videos per Row, Dim or Hide Watched Videos, and Auto Theater Mode are just some options.
    • Color Code Videos: Adds a configurable border around videos on the YouTube home page, with colors indicating video age and status.
  • Customizable
    • The settings panel allows for easy modification of the script and its functionality.
    • Includes buttons to export and import preferences for backup.
    • Color-Coded Interface to visually group related elements together
      • Green: Button One - NotebookLM
      • Blue: Button Two - ChatGPT
      • Red: Button Three - Download
      • White: Button Four - General Settings

Additional messages are displayed at the button's location in the YouTube header to indicate the status of the script.

Transcript Is Loading The transcript has to be loaded into the HTML before it can be used. This can take between a few milliseconds up to a couple of seconds—depending on the length of the video.
Transcript Not Available YouTube's API needs a few minutes to process a new video and generate its transcript.
Live videos don't have transcripts. Ensure the "Show transcript" button is available in the description.
Transcript Failed to Load If the transcript fails to load, the script will terminate. Reload the page to try again.

Changelog

  • 7.3.5: General bug fixes and improvements. Reorganized the settings panel for upcoming features. New Features: Hide YouTube brand text in the header; Hide video scrubber (red dot in progress bar); Display full titles; and Auto Theater Mode.
  • 7.3.1: fixed YouTube scrubber in progress bar
  • 7.3: New Features: Links in Header, Customize CSS, and Color Code Videos. Also optimized transcript loading and reset function.
  • 7.1: general bug fixes and improvements
  • 7.0: initial public release

License

This project is licensed under the AGPL-3.0 License. See the LICENSE file for details.