Skip to content

Image to Text (OCR)

Turn a photo of a document, a screenshot or a scanned PDF into editable text with optical character recognition (OCR). Recognition runs in your browser with the open-source Tesseract engine, so the file itself is never uploaded.

  • Free
  • No sign-up
  • Runs in your browser

All processing happens in your browser — files never leave your device. Your image or PDF is read on this device and is not uploaded. The OCR engine and language data download from the jsDelivr CDN on first use.

How it works

  1. Choose an image or a PDF of up to 20 pages.
  2. Select the document language, and a second language if the text mixes two.
  3. Choose Extract text, then review, copy or download the result as a .txt file.

Good to know: On first use the OCR engine and language data (about 2–15 MB per language) download from the jsDelivr CDN. Accuracy depends on image sharpness, lighting and layout; handwriting is not supported well.

How OCR reads your document

Optical character recognition finds lines of text in an image, splits them into words and characters, and matches each shape against a trained model of a language. Tesseract, the engine used here, was originally developed at HP and later maintained by Google, and is released under the Apache 2.0 licence. Its modern recognizer uses a neural network trained on each language's script, which is why choosing the correct language makes a large difference to accuracy.

For a scanned PDF, each page is first rendered to an image in your browser at twice its normal resolution and then recognized page by page. The output keeps line breaks but not columns, fonts or tables, so expect to tidy the layout afterwards.

Getting accurate results

Use a sharp, evenly lit image with the page filling most of the frame and the text horizontal. Around 300 dpi is ideal for scans; phone photos work well if you avoid shadows and blur. Very small text, decorative fonts, handwriting, low contrast and busy backgrounds reduce accuracy.

Always proofread important text such as names, numbers, dates and amounts. OCR commonly confuses similar shapes, for example 0 and O, 1 and l, or rn and m.

Frequently asked questions

Is my document uploaded?

No. The image or PDF is processed in your browser. Only the OCR engine and language files are downloaded, and they contain no information about your document.

Can it read handwriting?

Tesseract is designed for printed text. Neat block capitals may work, but cursive handwriting is usually not recognized reliably.

Why is the text in the wrong order?

Multi-column layouts, tables and text boxes can be read in an unexpected sequence. Crop the image to one column at a time for cleaner output.

Why is my OCR text full of strange characters?

The image is probably blurred, too small, or the wrong language is selected. Try a sharper photo, crop to the text and pick the right language.

Does OCR keep the layout of the page?

No. It returns plain text with line breaks. Tables, columns and fonts need to be rebuilt in your editor.

Related guides