OCR and advanced PDF tools
Run OCR on scanned PDFs, make them searchable, extract text and metadata, compare documents and search inside PDFs. Ten advanced tools — all browser-based, all private.
Tools in this category
What is OCR
OCR (Optical Character Recognition) is the process of reading the pixels of a scanned image and converting them into actual text. A scanned PDF contains pictures of pages, not text. OCR analyses those pictures and identifies the characters inside them, producing text that you can select, copy, search and index.
OmniPDF uses Tesseract.js, an open-source OCR engine compiled to WebAssembly. It runs entirely in your browser. No image or text is sent to a server. This is genuinely private OCR — unusual in a market where most OCR tools upload your document to a cloud service.
OCR accuracy honestly explained
Tesseract works well on clean, printed text at 300 DPI or higher. It works poorly on handwriting, stylised fonts, low-resolution photos, skewed pages, and text over complex backgrounds. If you scan a form on a flatbed scanner at 300 DPI in English, you will typically get 90%+ accuracy. If you photograph a handwritten note with your phone, expect much lower accuracy.
Language packs matter too. English is bundled by default. For French, German, Spanish and other languages, the corresponding traineddata file must be vendored under /assets/libraries/tesseract/. If a language pack is missing, OCR will fall back to English and produce incorrect output for that language.
Making a PDF searchable
A "searchable PDF" is a PDF that has an invisible text layer on top of the scanned image. The image stays as the visible content, but the text layer allows Ctrl+F and copy/paste to work. OmniPDF's Make PDF Searchable tool does this: it renders each page, runs OCR, then draws the recognised text back onto the page at very low opacity so it is searchable but not visible.
The alignment of the invisible text may not exactly match the visible image, so selecting a word sometimes selects a slightly offset rectangle. The text content is correct.
Comparing documents and searching
Compare Two PDFs extracts text from both files and highlights line-by-line differences. Useful for tracking edits between document versions.
Search Text in PDF finds a phrase across all pages and shows a snippet from each match with the page number.
What this category does not do
Translate PDF Text is not available on OmniPDF. Translation requires a network request to an external translation service, which would break the privacy model of the rest of the site. The tool page explains this honestly and points to alternatives.
Privacy
All OCR, comparison and search operations run in your browser. The document and the extracted text never leave your device.