How to Extract Text From an Image (Photos, Screenshots and Scans)

The fastest way to extract text from an image is to run it through OCR (optical character recognition), which turns the letters in a picture into text you can copy. Every phone and computer now has a free OCR tool built in, and a browser-based converter works on all of them. Pick the one below that matches your device.

1Extract text from an image online, for free

If you want one method that works the same on a phone, a work laptop or a Chromebook, use a browser-based converter. Our free Image to Text converter needs no app and no account.

1. Open urltotext.com/image-to-text.
2. Add your image in any of four ways:
   • Click Choose an image, or drop the file onto the box.
   • Copy a screenshot and press Ctrl + V (Cmd + V on a Mac) anywhere on the page.
   • On a phone, tap Take a photo to use the camera.
   • Switch to the Paste an image link tab, paste the image's URL, and click Convert.
3. The text appears in a few seconds. Click Copy, or Download to save it as a .txt file.

Open Options: language, line breaks before converting if you need either setting:

• Language of the text: pick from 19 languages, including Spanish, French, German, Russian, Arabic, Hindi, Chinese (Simplified and Traditional), Japanese and Korean. The right language makes a big difference to accuracy.
• Keep line breaks (receipts, lists, code): by default, wrapped lines are joined back into paragraphs, which is what you want for articles and documents. Turn this on to keep each line separate.

Good to know:

• It reads JPG, PNG, WebP, GIF, BMP and TIFF files up to 10 MB. HEIC photos from an iPhone aren't supported on a computer, so export them as JPG first (phones usually convert automatically when you upload).
• Small screenshots are enlarged and photos are turned upright automatically before reading, which helps with tiny text.
• Free use is limited to about 10 images every 10 minutes. Sign in to convert up to 10 images at once and download them all together.
• Your image is read in memory and thrown away as soon as the text is out. We don't keep a copy.
• It's built for printed and on-screen text. For handwriting, see method 8.

Have a scanned PDF instead of an image? Use PDF to Text, which reads every page, scanned or not. See our guide on how to extract text from a PDF.

2Use Google Lens in Chrome

Chrome on Windows, Mac and Chromebooks has Google Lens built in, so you can copy text from any image on a web page.

1. Right-click the image and choose Search with Google Lens.
2. A side panel opens with the image. Click Text at the top of the panel (some versions say Select text).
3. Click Select all text, or drag over just the part you need.
4. Click Copy text, then paste it anywhere.

Text that isn't an image (a video frame, a canvas, a chart): right-click an empty part of the page, choose Search with Google Lens, and drag a box around the text.

An image on your computer: go to google.com, click the camera icon (Search by image) in the search box, upload the file, and use the same Text option. On Windows, the Google app for desktop also does this: press Alt + Space, click the Lens icon, select part of your screen, and choose Copy as text.

Limitations: the image is sent to Google, so skip this for private documents. Lens sometimes reorders text from multi-column layouts.

3Use Live Text on iPhone and iPad

Live Text is built into iOS 15 and later (iPhone XS, iPhone XR or newer). It works in Photos, the Camera, screenshots, Safari and Quick Look, and it runs on the device, so nothing is uploaded.

First, check it's on: Settings›General›Language & Region, then turn on Live Text.

Copy text from a photo or screenshot

1. Open the photo in the Photos app.
2. Tap the Live Text button (a small frame with lines of text) at the bottom right of the photo. All the text is highlighted.
3. Touch and hold a word, then drag the grab points to choose what you want, or tap Select All.
4. Tap Copy, then paste it in Notes, Messages or any app.

For a screenshot, tap the thumbnail right after you take it, then touch and hold the text to select it. Or open it later in Photos and use the steps above.

Copy text straight from the Camera

1. Open Camera and point it at the text.
2. When a yellow frame appears around the text, tap the Live Text button in the corner of the viewfinder.
3. Select the text and tap Copy.

Tips: tap Translate in the Live Text menu to translate the text in place. Videos work too (iOS 16 and later): pause on the frame first. To type into a field from paper, tap the field, then tap the Scan Text button in the menu and point the camera at the text.

4Use Google Lens or Circle to Search on Android

Android phones read text with Google Lens, which lives in Google Photos, the Google app and Circle to Search.

From a photo or screenshot (Google Photos)

1. Open the image in Google Photos.
2. Tap Lens at the bottom of the screen.
3. Tap Select text (or tap a highlighted line), then Select all, or drag the handles.
4. Tap Copy text.

Google Photos on the web often shows a Copy text from image button at the top when it spots text in a photo.

From anything on screen (Circle to Search)

1. Touch and hold the Home button or the navigation handle at the bottom of the screen.
2. Circle or tap the text you want.
3. Tap Select text if needed, adjust the selection, and tap Copy.

Circle to Search works on Pixel 6 and later and many recent Samsung and other Android phones. Samsung Galaxy phones also have their own reader: in the Gallery app, tap the T icon that appears on photos with text.

5Use Snipping Tool or PowerToys on Windows

Snipping Tool (Windows 11)

Snipping Tool reads text on your PC itself, without uploading anything.

1. Press the Windows key + Shift + S, then drag around the text.
2. Click the notification to open the snip in Snipping Tool.
3. Click Text actions in the toolbar. The text is highlighted.
4. Click Copy all text, or drag over part of it, right-click and copy.

Quick redact blacks out email addresses and phone numbers before you share the image. Newer versions of Snipping Tool also add a Text extractor button to the capture bar that copies text without saving a screenshot. Microsoft has been rolling it out gradually, so you may not have it yet.

To read an image file you already have, open it in Snipping Tool with Ctrl + O and click Text actions.

PowerToys Text Extractor

If you copy text from the screen all day, Microsoft's free PowerToys adds a one-shortcut extractor:

1. Install PowerToys from the Microsoft Store.
2. Open PowerToys Settings, go to Text Extractor, and make sure it's turned on.
3. Press the Windows key + Shift + T, then drag over any text on screen.
4. The text is on your clipboard. Press Ctrl + V to paste.

Both tools use Windows' own OCR, which reads only languages whose OCR pack is installed. Add a language under Settings›Time & language›Language & region.

6Use Live Text on a Mac

Macs running macOS Monterey or later have Live Text in Preview, Photos, Quick Look and Safari. Turn it on under System Settings›General›Language & Region with the Live Text checkbox.

1. Open the image in Preview (or select it in Finder and press Space for Quick Look).
2. Move the pointer over the text until it becomes a text cursor, then drag to select. Press Cmd + A to select it all.
3. Press Cmd + C, or Control-click and choose Copy Text.

If you can't select anything in Preview, choose Tools›Text Selection. In Quick Look and Photos, click the Live Text button at the bottom right to highlight every word.

Screenshots: press Cmd + Shift + 4, drag over the text, then open the screenshot in Preview and copy as above.

Shortcuts: copy text from the screen in one step

1. Open the Shortcuts app and click + to make a new shortcut.
2. Add the Take Screenshot action, then Extract Text from Image, then Copy to Clipboard.
3. Name it, click the i (Shortcut Details) button and add a keyboard shortcut.

Run it and the text of your screen lands on the clipboard. Set Take Screenshot to Interactive to drag a selection instead of capturing the whole screen.

7Use Microsoft OneNote

OneNote's desktop app for Windows and Mac can read text from pictures in your notes:

1. Click Insert›Pictures and add the image (or paste it).
2. Right-click the picture and choose Copy Text from Picture.
3. Click where you want the text and press Ctrl + V.

If the option is grayed out, wait: Microsoft says recognition can take a while on complex images. Word doesn't read text from images directly, but it can convert a scanned PDF. Excel has Data›From Picture for turning a photo of a table into rows and columns.

8Ask ChatGPT, Claude or Gemini

AI chatbots read images well, and they're the best free option for handwriting, crumpled receipts and tables you want as a spreadsheet. Upload or paste the image with the + (or paperclip) button and ask for exactly what you want.

Use a prompt like this, so it copies rather than rewrites:

Transcribe all the text in this image exactly as written.
Keep the original line breaks. Don't fix spelling or grammar.
If a word is unreadable, write [illegible] instead of guessing.

For a table, ask: "Return this table as CSV with the same columns."

Watch out for hallucinations. Unlike OCR, a language model predicts text. It can:

• quietly "correct" spelling, names and numbers,
• fill in blurry words with something plausible,
• skip lines in long images or summarize instead of transcribing.

Check any number, date, name or price against the image before you rely on it. For long documents where every word matters, run OCR first and use the chatbot to tidy or summarize the result.

9Extract text from images with Python

For a folder of images, use Tesseract, the open-source OCR engine our converter runs on.

1. Install Tesseract: brew install tesseract on a Mac, sudo apt install tesseract-ocr on Ubuntu, or the UB Mannheim installer on Windows.
2. Install the Python packages: pip install pytesseract pillow.
3. Run a script like this:

from pathlib import Path

import pytesseract
from PIL import Image, ImageOps

for path in sorted(Path("images").glob("*.png")):
    image = ImageOps.grayscale(Image.open(path))
    if image.width < 1500:  # small text reads better enlarged
        image = image.resize((image.width * 2, image.height * 2))
    text = pytesseract.image_to_string(image, lang="eng")
    path.with_suffix(".txt").write_text(text, encoding="utf-8")
    print(path.name, len(text.split()), "words")

Tips:

• Add languages with sudo apt install tesseract-ocr-fra (or brew install tesseract-lang), then use lang="eng+fra".
• Pass config="--psm 6" for one uniform block of text, such as a receipt, or --psm 7 for a single line.
• Tesseract expects clean, straight, printed text. For photos taken at an angle, straighten and crop first (OpenCV helps), or try EasyOCR or PaddleOCR, which handle scene text better.

Handwriting, tables and other languages

• Handwriting: Google Lens, Apple Live Text and the AI chatbots do best with neat handwriting. Classic OCR, including Tesseract, is built for print and struggles with cursive.
• Tables: turn on Keep line breaks in Image to Text so each row stays on its own line, use Excel's From Picture, or ask a chatbot for CSV.
• Other languages: always set the language. Our converter, Tesseract and Windows OCR all read much better when they know the script. Lens and Live Text detect it automatically.
• Text on a web page: don't screenshot it. Copy the real text with URL to Text, which is exact and keeps links.

How to get more accurate results

OCR is only as good as the image you feed it:

• Resolution: letters should be at least 20 pixels tall. Scan at 300 DPI. For screenshots, zoom in (Ctrl + +) before capturing.
• Contrast: dark text on a plain light background works best. Avoid shadows, glare and colored backgrounds.
• Straight and flat: shoot from directly above, flatten curled pages, and rotate or deskew crooked scans.
• Crop: cut away everything that isn't text, especially photos and logos next to it.
• File type: save screenshots as PNG. Heavy JPG compression blurs letter edges.
• One column at a time: for multi-column pages, crop each column separately to keep the reading order.

Frequently asked questions

How do I extract text from an image for free?

Upload or paste it into the free Image to Text converter and click Copy. You can also use the tools already on your device: Live Text on iPhone and Mac, Google Lens on Android and in Chrome, and Snipping Tool's Text actions on Windows.

How do I copy text from a screenshot?

On Windows, open the screenshot in Snipping Tool and click Text actions, then Copy all text. On iPhone, open it in Photos and tap the Live Text button. On a Mac, open it in Preview and drag over the text. On any device, press Ctrl + V on the Image to Text page to paste the screenshot straight in.

Can Google convert an image to text?

Yes. Google Lens reads text in images in Chrome, Google Photos and the Google app. Google Drive can too: right-click an image of 2 MB or less and choose Open with›Google Docs, and the text appears below the picture.

Can OCR read handwriting?

Sometimes. Neat, printed handwriting often works in Google Lens, Apple Live Text and AI chatbots. Cursive and messy notes are hit and miss everywhere, so check the result carefully.

Why is my extracted text full of mistakes?

Usually the image is too small, blurry, crooked or low contrast, or the wrong language is set. Crop to the text, use a sharper or larger image, and pick the right language. Decorative fonts and text over photos also confuse OCR.

Is it safe to upload images to an online OCR tool?

It depends on the tool. Our Image to Text converter reads your image in memory and discards it right away. For confidential documents, Live Text, Snipping Tool and PowerToys run entirely on your device.

Can I extract text from a video?

Pause the video and use Live Text (iPhone, iPad, Mac), Circle to Search (Android), or PowerToys Text Extractor (Windows). For spoken words, use a transcript tool instead.

Which method should you use?

• Any device, no setup: Image to Text.
• iPhone, iPad, Mac: Live Text.
• Android: Google Lens or Circle to Search.
• Windows: Snipping Tool, or PowerToys for frequent use.
• Handwriting or tables: a chatbot, then double-check it.
• Hundreds of images: Python and Tesseract.