How to Copy Text from a Video or a YouTube Frame

OCR guide · 4 min read

You are watching a tutorial and the exact command you need is right there on screen. You try to select it, drag your cursor across the words, and nothing happens. The text refuses to highlight. That is not a glitch, and it is not your browser. A video is simply a sequence of images, and text painted into those images is not selectable text at all.

The good news is that grabbing those words is easy once you understand what is actually happening. This guide explains why video text cannot be copied directly, then walks through a reliable way to pull it out using on-device OCR.

Why you can't select text in a video

When you highlight text on a normal web page, your browser knows where each character sits because the page contains real, encoded characters. A video frame contains none of that. Every frame is a grid of colored pixels, and a word like "npm install" is just an arrangement of light and dark dots that looks like letters to you. The player has no idea there is text in there, so there is nothing to select and nothing to copy.

The same is true for burned-in captions, slide decks in a webinar recording, code in a screencast, and a phone number flashed on screen. To a computer it is all picture, not writing. To get usable text back, you need to convert those pixels into characters. That conversion is called OCR (optical character recognition), and it is exactly the step that turns a paused frame into text you can paste.

The basic technique: pause, then OCR the frame

The method works for any video on any site, because you are reading the screen rather than the video file:

1. Pause on the frame you want. Scrub to the moment the text is fully visible and stop there.
2. Run OCR on that region of the screen. Point an OCR tool at the part of the frame holding the text.
3. Get the characters back. The tool reads the pixels and hands you real, editable text.

A browser extension is a convenient place to do this, since the video is already in a tab. Textquill runs OCR on your own device, so the frame never leaves your computer.

Option A: Select just the area you need (Alt+Shift+S)

When the text is one part of the frame, such as a line of code or a caption, grab only that region:

• Pause the video on the frame.
• Press Alt+Shift+S to start area select.
• Drag a box around the text you want.
• Release, and the recognized text appears ready to copy.

This is the fastest choice most of the time, because you skip everything around the text and give the OCR engine a clean, tight crop to work with.

Option B: Capture the visible tab

If you want everything on screen, for example a full lecture slide or a table, capture the whole visible tab instead and let the tool read the entire frame. You can then keep the parts you need. This is handy when there are several blocks of text and drawing one box would miss some of them.

Where this actually comes in handy

• Lecture slides. Pull bullet points and definitions from a recorded class instead of retyping them into your notes.
• Tutorial code. Copy a command or snippet shown in a screencast so you can run it, rather than squinting and guessing at each character.
• Webinar details. Grab an email address, a URL, or a discount code that appears on a slide for a few seconds.
• Captions and subtitles. Save a quote or a line of dialogue from burned-in captions that you cannot otherwise select.
• A product name in a review. Read the exact model number or brand shown in a hands-on video so you can search for it later.

Tips for a clean grab

OCR accuracy depends almost entirely on how clear the frame is. A few habits make a real difference:

• Pause on a sharp frame. Nudge forward or back a frame if the text looks smeared. Motion blur is the most common reason characters come out wrong.
• Go full screen first. A larger video means larger text and more pixels per letter, which OCR reads more reliably.
• Prefer higher resolution. Bump the player quality up to 1080p or better before you pause. Low-quality streams blur small text.
• Frame tightly. With area select, box only the text. Extra background and graphics give the engine less to trip over.

What to do with the text after

Once the frame becomes real text, you can treat it like anything else you typed. Copy it straight into your notes, an editor, or a chat. Have it read aloud to check a long string such as a URL or a code without staring at the screen. Or export it as TXT or Markdown if you are collecting several grabs from one video. Because Textquill keeps a searchable history, you can also come back later and find that snippet by a word you remember from it.

One more reason to do this inside the browser: many "screenshot to text" websites ask you to upload the image to their servers. Reading the frame on your own device with on-device OCR keeps that slide, caption, or contact detail private, and it works even when you are offline.

FAQ