Upload a scanned PDF — the OCR engine reads every page and produces a searchable, text-selectable PDF. All processing happens in your browser.
🔒 No uploads — fully local 🌍 10 languages 🔍 Ctrl+F after OCR 🆓 Always free
What You Get
Searchable PDF — same visual appearance as the original, but now you can press Ctrl+F to search inside it, select and copy text, and have it indexed by your document management system. Plain TXT — all recognized text exported as a plain text file, useful for further processing, importing into spreadsheets, or feeding to AI tools.
Use Cases
Making old scanned contracts searchable before uploading to a document management system
Extracting text from photographed receipts for expense reports
Making scanned academic papers accessible and full-text searchable
Converting paper forms that were scanned into usable data
Frequently Asked Questions
What is OCR and why do I need it?
OCR (Optical Character Recognition) converts images of text — like a scanned document or a photographed page — into actual machine-readable text. Without OCR, a scanned PDF is just a collection of images: you can't search it, copy text from it, or have it read by a screen reader. After OCR, the PDF becomes fully searchable and accessible.
What types of PDFs need OCR?
Any PDF where the content is an image rather than text. This includes: documents scanned on a physical scanner, photographed documents (photo of a paper contract, receipt, etc.), PDF exports from older fax machines, and PDFs where text is stored as images (sometimes done to prevent copying). If you open the PDF and can't select or copy any text, it needs OCR.
How long does OCR take?
OCR time depends on the number of pages and the render quality setting. A 10-page document at 2× quality (recommended) typically takes 30–90 seconds on a modern laptop. At 1× quality, it's faster but may miss small text. At 3×, it's slower but best for documents with small print or handwriting.
What languages are supported?
English, French, German, Spanish, Portuguese, Italian, Chinese (Simplified), Japanese, Arabic, and Hindi. The OCR engine (Tesseract.js) downloads the language data (~5 MB per language) the first time you use it — it's cached in your browser for future use.