Make scanned PDFs searchable
Run OCR on scanned PDFs to make them searchable, copyable, and indexable. 100+ languages including Arabic, Chinese, Spanish, French, and German. Powered by Tesseract + OCRmyPDF.
Drag & drop a file
or browse from your computer · max 100 MB
Pick the language(s) of text in your scan. The output is a searchable PDF.
OCR (Optical Character Recognition) turns a scanned PDF — which is technically just a stack of images — into a real searchable, copyable, indexable document. Without OCR, you can't Ctrl+F a scanned contract, can't quote from a photographed receipt, and Google can't index the content. PDFOnly's OCR engine uses Tesseract via OCRmyPDF with deskew and auto-rotation pre-processing built in. Output is a standard PDF that looks identical to your scan but has an invisible text layer behind every word — fully searchable in any PDF reader, fully indexable by search engines.
How to ocr pdf step by step
- 1
Upload the scanned PDF
Any PDF — typically scans from a printer, phone camera, or document scanner. Files up to 100 MB free, 200 MB Pro. We support most scan formats including grayscale, color, and bitonal (black and white).
- 2
Pick the language(s)
Default is English. For multi-language documents, you can pick multiple — e.g. 'eng+ara' for English-and-Arabic mixed pages. Auto-detect mode tries to identify the language from the page content, but explicit language picking is more accurate.
- 3
We deskew, rotate, and OCR
Pre-processing detects and corrects skew (pages scanned at slight angles) and rotation (pages scanned sideways). Then Tesseract recognizes characters page by page and adds an invisible text layer.
- 4
Download a searchable PDF
The output looks visually identical to your scan but has a hidden text layer aligned to every word. Open it in any PDF reader and Ctrl+F searches actually find things now.
Why ocr pdf on PDFOnly
Real OCR pipeline
We use OCRmyPDF (which wraps Tesseract + Ghostscript + qpdf) — the same toolchain used by enterprise scanning solutions. Most online OCR tools are wrappers around outdated engines.
100+ languages
Arabic, Chinese (simplified and traditional), Japanese, Korean, Hebrew, Hindi, Thai, Vietnamese, plus all major European languages. Most free tools only support English well.
Skip-text mode
If your PDF is partly searchable already, we don't re-OCR those pages — saves time and avoids degrading existing text quality. Most tools blindly re-OCR everything.
What people use ocr pdf for
A few common scenarios. If your workflow looks like one of these, this tool is a good fit.
Make scanned legal exhibits Ctrl+F-searchable
Lawyers reviewing 500-page exhibit binders save hours when they can search a witness name across the whole pile instead of flipping page by page.
Index a scanned book or thesis for citation
Researchers can quote directly from scanned older texts after OCR — no more retyping passages by hand.
Make a receipt archive searchable
Snapshots of receipts become searchable by vendor name, date, or item — turning a shoebox into a queryable database.
Convert non-Latin scripts
Arabic, Chinese, Japanese, and Russian OCR have historically been weak in free tools. We tune Tesseract specifically for these scripts and get 90%+ accuracy on clean scans.
What you get
- 100+ languages including Arabic, Mandarin, Spanish, French, German, Russian, Hindi
- Auto-deskew and auto-rotate before OCR for sharper recognition
- Searchable PDF output looks identical to the scan — text layer is invisible
- Skip-text mode: leaves already-searchable PDFs untouched, only OCRs the rest
- 95%+ accuracy on clean scans, 80-90% on faded or low-resolution
- Files auto-deleted in one hour, never used to train AI
- Free for files up to 100 MB
Frequently asked questions
How accurate is the OCR?
On clean, modern scans at 300 DPI: 95-99% character accuracy in English and most Latin-script languages. Arabic and CJK languages typically 90-95%. Faded, low-resolution, or skewed scans drop sharply — re-scan at 300 DPI if you need higher accuracy.
Will the output PDF look different from my scan?
No. The text layer is invisible — it sits behind the original image so the document looks pixel-identical. Only difference: now Ctrl+F works and you can copy text out.
Can I OCR a multi-language document?
Yes. Pick multiple languages in the options panel (e.g. English + Arabic for a bilingual contract). Tesseract handles mixed-language pages well when both languages are specified up front.
What about handwriting?
Tesseract is optimized for printed text. Handwriting recognition (HTR) is a different model — accuracy on handwriting drops to 30-60% with our engine. For handwriting, dedicated tools like Google Document AI work better.
Will OCR make my PDF larger?
Yes, but only slightly — adding the text layer increases file size by ~1-3%. Run our Compress PDF tool afterwards if size matters; the searchable layer survives compression.
Can I just extract the text without keeping the original PDF?
Use our PDF-to-Word tool instead — it runs OCR for scans and outputs editable .docx. OCR PDF keeps the original visual layout; PDF-to-Word gives you raw editable text.
What languages do you support?
100+: English, Arabic, Spanish, French, German, Russian, Italian, Portuguese, Dutch, Swedish, Polish, Czech, Greek, Turkish, Hebrew, Hindi, Bengali, Tamil, Chinese Simplified, Chinese Traditional, Japanese, Korean, Vietnamese, Thai, Indonesian, and more. See full list in the language picker.
Ready to ocr pdf?
Free to use for the basics. Files are auto-deleted within an hour and never used to train AI.
Open OCR PDF