OCR PDF
Read text out of scanned PDFs and photos with real optical character recognition — running entirely in your browser.
Runs in your browser — nothing is uploaded
Click to upload or drag and drop
PDF, PNG, JPG, WEBP, BMP or TIFF
Choosing the right language makes a large difference to accuracy.
How OCR PDF Works
This tool runs Tesseract, the open-source OCR engine, compiled to WebAssembly so it executes inside your browser tab. Each page of your PDF is rendered to an image, analysed character by character, and turned back into selectable text.
Two outputs. You get a searchable PDF — your original pages with an invisible text layer behind them, so you can select, copy and Ctrl+F the contents — and a plain .txt file of everything that was recognised.
Works on scans and photos. Unlike a plain text extractor, this reads documents that contain no text at all: scanned contracts, photographed receipts, screenshots. Upload a PDF or an image file.
Privacy. Your document never leaves your device — there is no upload and no server. The recognition engine and the language model are downloaded once from a public CDN and then cached by your browser; those downloads contain no data of yours.
Tips for accuracy. Pick the language that matches your document. Straight, well-lit, high-resolution scans read far better than skewed phone photos. Large documents take time — OCR is doing real work on your CPU, roughly a few seconds per page.
Frequently asked questions
Does this do real OCR on scanned documents?
Yes. It runs Tesseract, the open-source OCR engine, compiled to WebAssembly so it executes in your browser. It reads documents containing no text layer at all — scanned contracts, photographed receipts, screenshots — by recognising the shapes of the characters.
What is a searchable PDF?
It is your original scanned pages with an invisible text layer positioned behind them. The document looks identical, but you can select, copy and search the text with Ctrl+F. This tool produces one alongside a plain .txt file.
Which languages are supported?
Twelve, including English, Spanish, French, German, Portuguese, Italian, Dutch, Russian, Arabic, Hindi, Chinese (Simplified) and Japanese. Choosing the language that matches your document makes a substantial difference to accuracy.
How can I improve OCR accuracy?
Select the correct language, and start from the sharpest scan you have. Straight, well-lit, high-resolution pages read far better than angled phone photos. Text over busy backgrounds or in unusual decorative fonts is always harder.
Is my scanned document uploaded?
No. Your document is processed entirely on your device. The recognition engine and language model are downloaded once from a public CDN and cached by your browser — those downloads contain no data of yours. This is the key difference from other online OCR services, which upload your documents to their servers.