How to Turn a YouTube Video into Images

Last updated: 7 November 2026 · By the FrameCap team

Use FrameCap to capture a picture every few seconds. You get every moment of the part you pick as images, each with its time: a PDF for AI chatbots, or PNG or JPG pictures for AI agents. It is free in Chrome.

Pick how often

A talk with slides: every 10 seconds.
A tutorial: every 2 to 5 seconds.
Code typed live: every 1 second.

Pick what to keep

Pictures only gives you the images. Pictures and words adds the transcript, so an AI can match what was said to what was on screen.

Pick the part

Capture the whole video, start from where you are, or set your own start and end. A short part keeps the result easy to read.

Fewer repeated images

When the screen stays the same for a while, FrameCap Pro's Smart frames skips the pictures that would look the same.

What to do with the images

Ask your AI to list the steps, find the moment something appears, or write notes with times. FrameCap Pro can also save notes for Obsidian or Notion.

FAQ

Can I turn a 2-hour lecture into images?

Yes. The free version takes up to 40 minutes at a time and 2 hours a day, so capture it in parts. FrameCap Pro takes up to 3 hours at once.

Do I get the words too?

Yes, if the video has captions. Automatic captions are fine. Pick Pictures and words.

Which is better for AI, the PDF or the ZIP?

The PDF is made for AI chatbots like ChatGPT and Claude. The ZIP is made for AI agents like Claude Code, Codex and Cursor.