Image to text (OCR)

Choose a photo, screenshot or scan, pick the language and press Extract text: Tesseract OCR reads the text on your device (13 languages, including French, Spanish, German, Arabic, Hindi and Chinese), then you can copy it or download it as a .txt file. Nothing is uploaded.

  • ๐Ÿ”’ Runs on your device
  • โค’ 0 bytes uploaded
  • โœ“ Free, no sign-up, no watermark
  • โฌ‡ 7 MB engine, downloaded once
Loading the toolโ€ฆ

How to use Image to Text (OCR)

  1. Choose or drop one or more images (JPG, PNG, WebP, HEICโ€ฆ).
  2. Keep "Clean up the image first" on for phone photos.
  3. Press Extract text and wait for the engine (downloaded once, about 7 MB).
  4. Copy the text or download it as a .txt file.

About Image to Text (OCR)

Nolo runs Tesseract, the open-source OCR engine originally developed at HP and later Google, compiled to WebAssembly (tesseract.js) with its LSTM neural-network language models. Recognition runs in a background worker so the page stays responsive.

Results are not perfect: always check names, amounts and numbers. Average confidence is shown so you know how sure the engine was.

Is it private?

Yes. Image to Text (OCR) runs entirely in your web browser using your device's own processor. Your files never leave your device: they are not uploaded to Nolo or anyone else, and the page's security policy only allows it to talk to nolotools.com, which accepts no files. The site sends only two small things: the words typed in the optional search box, to find a tool, and, the first time you use a tool on a page, that tool's name for an anonymous daily count of tool use (nothing about you, your files or what you typed). Results are kept encrypted on your device for 60 minutes in Your files, then deleted automatically, or straight away with Delete all. How Nolo works.

Questions and answers

Are my images uploaded?

No. Everything happens in your browser on your own device. Your images are never uploaded to Nolo or anyone else, and the tool keeps working with the internet switched off.

Which languages can it read?

English, French, Spanish, Portuguese, German, Italian, Russian, Arabic, Hindi, Chinese (simplified and traditional), Japanese and Korean. Pick the language of the text; each language (1 to 3 MB) is downloaded from Nolo the first time you use it.

Can it read handwriting?

Poorly. Tesseract is built for printed text. Neat block capitals sometimes work; cursive usually does not.

How do I get better results?

Use a sharp, well-lit photo taken straight on, crop to just the text, and make sure the text is at least about 20 pixels tall.

Does it work offline?

Yes, after the first use. The OCR engine (about 7 MB) and each language you use (1 to 3 MB) are downloaded once from Nolo and then stored by your browser.

Can I OCR a PDF?

Yes, with the PDF tools: OCR PDF makes a scanned PDF searchable, and PDF to Text and PDF to Word read scanned pages with OCR.