Convert a YouTube video to PDF, slides included

PDF is the one format everything accepts. Your AI reads it, your phone opens it, your printer prints it, your teammate cannot mess up the formatting. So when I want a video to become something permanent, PDF is what I aim for, and I learned the hard way that the words alone are not enough. The first time I turned a lecture into a plain transcript PDF, I opened it a month later and half of it made no sense, because the teacher kept saying "as you can see here" about slides I could no longer see.

A good video PDF needs both layers. The words, timestamped so you can jump back to the exact moment. And the screen, captured as images, because in a tutorial the screen carries the half of the meaning that never gets spoken.

The free trick, and where it stops

If you just need words on paper today, do this:

1. Open the video on YouTube and click the three dots under it, then Show transcript.
2. Copy the whole panel and paste it into a Google Doc.
3. File, Download, PDF. Done.

It works, and it costs nothing. But you inherit every weakness of auto captions: missing videos, mangled technical words, no punctuation worth trusting. You get zero images, no structure, and cleaning the timestamp clutter by hand takes longer than you expect. I still use this trick for short interview clips. For anything taught on a screen, it throws away the screen.

The proper way: link in, PDF out

VideoDoc was built for exactly this output. You paste a YouTube link, a single video or a whole playlist merged into one document, choose your quality, and press Start. Everything runs on your own machine: the download comes over your home connection, the transcription runs on your CPU or graphics card, and the screenshot engine watches the picture as it goes.

That screenshot engine is the part I am most proud of. When a video mixes a person and a screen share, it follows the screen, not the face. The slides, the code editor, the whiteboard, captured at up to full HD with Best quality so the text inside the image stays sharp enough for your AI to read. A speaker's face gets captured only when the face is genuinely all there is.

What the finished PDF looks like inside

• The video title, channel, link and length at the top, so the source is never lost.
• The timestamped transcript, broken into readable blocks rather than caption fragments.
• The on-screen captures placed inline, each one sitting next to the words that were spoken over it.
• A matching Markdown file beside the PDF, if you prefer notes apps like Obsidian or Notion.

Because the PDF is a normal document, ChatGPT, Claude, Gemini and NotebookLM all read it directly. Ask for a summary that keeps the timestamps, ask what the third slide meant, ask it to quiz you. The AI finally has the whole video in a form it understands.

Quick questions

Pick the video you reference most often and make it a PDF tonight. Searchable beats rewatchable, every single time.