Image to Text (OCR)
Extract English text from screenshots and photos with OCR.
Files are processed entirely in your browser — they're never uploaded to a server.
Reads English text. The first scan downloads the reading engine and English language data (up to about 7 MB), so it takes a little longer; later scans start straight away.
Copy text out of a picture
Image to Text reads the printed words in a screenshot, photo or scanned image and turns them into text you can edit, copy or save. It uses Tesseract, an open-source OCR (optical character recognition) engine, running in your browser.
It reads English only. The engine and the English language data are downloaded from this site the first time you click Extract text, so that first scan takes longer. Your image stays on your device throughout.
How to extract text from an image
- Click the upload area, drag an image onto it, or paste a screenshot with Ctrl+V. It reads PNG, JPG or WebP images.
- Check the preview to make sure it’s the right image.
- Click Extract text. The progress bar shows the engine loading on first use, then the reading progress.
- Review the result in the Extracted text box. You can type in it to fix any misread words.
- Click Copy text, or Download .txt to save it as a plain text file named after your image.
Reading the output and the confidence score
Line breaks roughly follow the lines in the image, with long runs of blank lines tidied away. Layout such as columns and tables isn’t kept — you get plain text.
The Confidence figure is the engine’s average certainty about the words it recognised. A high number usually means few errors and a low one means proofread closely, but it isn’t a guarantee: always check names, numbers and punctuation.
Things people use OCR for
- Unselectable text: error messages, slides shared in a video call, or text inside an app screenshot.
- Printed pages: photograph a letter, recipe or book page and get the words without retyping.
- Scanned PDFs: turn the pages into images with PDF to Image, then read them here. For PDFs that already have selectable text, PDF to Word is quicker and keeps paragraphs.
- Labels and receipts: copy reference numbers or addresses, then double-check them against the original.
How to get more accurate text
- Crop the image to just the text you need — logos, photos and busy backgrounds confuse the engine.
- Keep text horizontal. Straighten tilted photos before uploading.
- Use even lighting with no shadows or glare, and make sure the photo is in focus.
- Bigger is better: zoom in before taking a screenshot of small text.
- Screenshots of on-screen text usually read better than photos of paper.
Image to Text (OCR): common questions
Can it read handwriting?
It’s built for printed and on-screen text. Handwriting usually comes out with many errors or not at all.
Which languages does it support?
English only. Other languages written in the Latin alphabet may come out partly readable, but accented letters are often misread and other scripts aren’t recognised.
Can I upload a PDF?
No, it takes images only. Convert the pages with PDF to Image first, or browse the other PDF tools if your PDF has real text.
Why is the first scan slow?
The reading engine and English data are several megabytes and download on first use. After that, scans start straight away.
Why does it say no text was found?
The engine couldn’t pick out any words — common with blurry, dark, tiny or sideways text. Try a sharper, well-lit image with the text level.
Is my image sent to a server for reading?
No. The OCR engine is loaded from this site and runs in your browser tab; no outside service sees your image or the extracted text.