How to convert an image to text with OCR
- 1 Drop an image into the box, press browse, or paste a screenshot.
- 2 Wait while the text is read. The first time, the reader takes a few seconds to load.
- 3 Check and correct the text in the box if needed.
- 4 Press Copy text or Download .txt.
Getting accurate results from OCR
OCR (optical character recognition) looks at the shapes in a picture and guesses which letters they are. It is very accurate on clear printed text and less reliable on everything else.
- Best results: screenshots and flat, well-lit photos of printed pages, with text at least 20 pixels tall.
- Straighten first: a tilted or sideways page lowers accuracy, so fix it with Rotate Image and crop to the text area.
- Handwriting and decorative fonts are often misread; check names and numbers by eye.
The engine reads English text. If your text is inside a normal PDF, PDF to Text copies it exactly with no guessing. Once you have the words, the Word Counter counts them.
Features
Accurate OCR
Uses the Tesseract engine trained on English text.
Screenshots and photos
Works with screenshots, phone photos and scans.
Confidence score
Shows how sure the reader is about the result.
Private
The image is read on your device.
When to use it
Copying from screenshots
Grab text you cannot select. Crop to just the text area with Crop Image for better accuracy.
Digitising notes
Turn printed pages into editable text. Turn a sideways photo of a page with Rotate Image first.
Receipts and cards
Copy details from photos. iPhone photos in HEIC can be converted with HEIC to JPG.
Quotes
Copy text from an image to share it. Very large photos load faster after Resize Image.
Related Image tasks
Frequently asked questions
Are my images uploaded to a server?
No. Your images are processed by your own browser and never leave your device, so private photos, ID cards and documents stay private. Once the page has loaded, the tool even works offline.
Which languages are supported?
English is supported now. Text in other scripts, such as Urdu or Hindi, will not be read correctly.
How accurate is it?
For clear screenshots and printed text, it is usually very accurate. Blurry, angled or handwritten text gives poorer results.
Can it read handwriting?
Neat block capitals may work, but cursive handwriting is not read reliably.
How do I get the best result?
Use a sharp, well-lit image taken straight on, with dark text on a light background. Crop out anything that is not text.
Why does the first image take longer?
The text reader and its English language data load the first time, about 9 MB. After that, images are read much faster.
Can I extract text from a PDF?
Use our PDF to Text tool for PDFs that contain real text.
Is it free?
Yes. There is no sign-up, no watermark and no daily limit.
Can I paste a screenshot?
Yes. Take a screenshot and press Ctrl+V or Cmd+V on this page, and the text is read straight away.
Can I edit the text before copying?
Yes. The text box is editable, so you can fix any mistakes before copying or downloading.
How this tool works. These image tools run entirely in your browser using the Canvas API built into every modern browser, the open-source libheif decoder for iPhone HEIC photos and Tesseract, a widely used open-source OCR engine, for reading text. Each tool is tested with real phone photos, screenshots and scans before it goes live, and nothing you open is sent to a server.
Built and reviewed by Hamza Farooq, founder and lead developer
· Last updated
· How we check our tools
· Report a problem