How to make a scanned PDF searchable
A scanned PDF is just photographs of pages — Ctrl+F finds nothing, text can't be copied, and screen readers are blind to it. OCR (optical character recognition) fixes that by adding an invisible text layer over each page image. You can do it free, in your browser, with no page limits.
Step by step
- Clean the scan first if it's rough. Skewed, gray, low-contrast scans OCR badly. Run crooked scans through Clean Up Scan first — auto-straighten plus text thickening measurably improves recognition.
- Drop the PDF into the OCR tool. Images work too (JPG, PNG, TIFF — even multi-page TIFFs). Everything is recognized locally by a WebAssembly build of Tesseract; nothing uploads, and there's no 10-page cap like the freemium sites.
- Pick the document's language. This matters more than people expect — the language model drives recognition. 15 languages are available; the model downloads once and is cached.
- Choose “Searchable PDF” and run. Each page keeps its original look, with the recognized text laid invisibly on top — selectable, copyable, searchable. Or choose plain text output if you just want the words.
Worth knowing
Recognition quality tracks scan quality: 300 DPI scans OCR beautifully, phone photos of documents are hit-and-miss (flatten the page, avoid shadows). For phone captures, Clean Up Scan's auto-deskew is the difference between 90% and 99% accuracy.
Frequently asked questions
How accurate is it?
On a clean printed scan, expect near-perfect results in the main languages. Handwriting is not supported — that's a genuinely different problem that even the big cloud OCRs handle inconsistently.
Is there really no page limit?
Really. A 400-page scanned book works — it just takes a while, since recognition runs on your device. Progress is shown page by page.