Skip to content
0 Bytesuploaded
Language
Explore Tools

Free Offline PDF & Image OCR

Drop a scanned PDF, a photo of a document or a screenshot. Each page is rendered at about 300 DPI and read by Tesseract, the open-source OCR engine, compiled to WebAssembly and running entirely in your browser. Copy the text or download it as a .txt file.

Quick answer

How does offline OCR work?

Tesseract.js, a WebAssembly build of the Tesseract LSTM engine, recognises text in scanned PDF pages and images inside your browser. The English and German language data is served from this site, so recognition keeps working offline once loaded. You get a confidence score per page, and 0 bytes are uploaded.

How it works

  1. 01

    Drop scanned PDFs or images: PDF, JPG, PNG, HEIC or WebP.

  2. 02

    Choose the document language: English, German or both.

  3. 03

    Press Recognise text, then copy it or download a .txt file.

OCR a PDF document online without uploading it

Most online OCR services upload your scan to their servers, which is a problem for contracts, medical letters and ID documents. Here the OCR engine and its language data are downloaded once from this site, then everything runs locally. You can disconnect from the internet and it keeps working.

PDF to text with OCR

Multi-page PDFs are processed page by page with live progress. The output keeps page breaks so you can find your place, and the confidence score tells you how clean the scan was.

Tips for better OCR results

Straight, evenly lit scans work best. Choose the right language so umlauts and ligatures are recognised, crop away dark borders before scanning, and prefer the original PDF over a photo of a screen.

Frequently asked questions

How do I OCR a PDF?

Drop a scanned PDF, choose the document language and press Recognise text. Each page is read with live progress, then you copy the text or download it as a .txt file.

What is PDF OCR?

OCR (optical character recognition) finds the letters inside the page images of a scanned PDF and turns them into real text you can copy, search and edit.

Which languages are supported?

English and German, alone or together for mixed documents. The language data is hosted on this site, so recognition works offline once the page has loaded.

How accurate is the text recognition?

Clean, straight scans usually reach 90–99%, and each result shows a confidence score. Low-resolution photos, skewed pages and handwriting reduce accuracy.

Are my scans uploaded?

No. The file is processed in your browser's memory by a Web Worker on your device and is never sent to a server. Once the page has loaded you can switch off Wi-Fi and it still works.

Related tools