Highly accurate on clean 300 DPI scans of printed text — near-perfect for English and major languages. Accuracy drops on phone-camera shots, skewed pages, low-resolution scans, handwriting, and small fonts. PDFWix will report confidence per page at launch so you know what to trust.

Will there be a file size limit?

We're not planning a hard cap, but very large or high-resolution scans will be slower because OCR is computationally heavy. Most everyday documents (under a few hundred pages) will process in well under a minute.

Will my file be uploaded?

Where device performance allows, OCR will run entirely in your browser. For very large files where browser-side OCR would be impractical, processing will fall back to server-side over HTTPS, in memory only — never written to disk.

OCR a scanned PDF

Make scanned PDFs searchable with high-accuracy OCR in 100+ languages.

Why use this

Searchable, selectable, accessible: OCR adds a real text layer so Ctrl+F works, copy-paste works, and screen readers can read the document — the three things scanned PDFs can't do without it.
Visual preserved exactly: OCR adds an invisible text layer beneath the image — the page still looks like the original scan, just with hidden searchable text underneath.
100+ languages on the roadmap: Latin, Cyrillic, Greek, Arabic, Chinese, Japanese, Korean and many more will be supported at launch. Multi-language pages can mix scripts.
Honest about limits: OCR isn't perfect. Accuracy depends heavily on scan quality — clean 300 DPI scans recover near-perfectly, phone-camera shots at angle recover less. PDFWix shows confidence per page so you know what…
Browser-side where possible: Privacy-sensitive scans (medical records, financials) will be processed locally in your browser whenever device performance allows. No file leaves your device unnecessarily.

Use cases

Making a scanned book PDF searchable so you can Ctrl+F for a quote
Adding a text layer to scanned legal exhibits so they're indexable in a case management system
Recovering text from a scanned receipt for an expense report
Making a scanned medical record accessible to a screen reader for a vision-impaired patient
Turning a 1990s-era scanned thesis into a searchable archive for a library

How it works

Open the OCR tool — When the tool launches, you'll click "Select PDF file" or drag a scanned PDF into the upload box.
Pick the language(s) — Choose the one or two languages actually present in the document. Single-language OCR is faster and more accurate than running every language at once.
Click "Run OCR" — PDFWix will analyze each page, recognise the text, and inject an invisible searchable layer beneath the original image — preserving the visual exactly while making the content searchable.
Download the searchable PDF — Save the OCR'd file. Open in any reader and Ctrl+F should now find words you can see in the document. The original visual appearance is unchanged.

Frequently asked questions

Is OCR available now?

Not yet — it's on our launch list and arriving soon. Join the waitlist on the page and we'll email you the moment it's live. In the meantime, Adobe Acrobat Pro and Google Drive's 'open as Google Docs' both run a passable OCR pass on scanned PDFs.

Which languages will be supported?

We're targeting 100+ at launch including English, Spanish, French, German, Italian, Portuguese, Dutch, Russian, Chinese (simplified and traditional), Japanese, Korean, Arabic, Hindi, Bengali and more. Multi-language pages will be allowed.

Will the visual look the same after OCR?

Yes. OCR adds a hidden searchable text layer underneath the original image — the page looks identical to the source scan. You'll only notice the difference when you Ctrl+F or try to copy text.