OCR PDF — Make Scanned Documents Searchable
Turn scanned PDFs into searchable, selectable, copy-paste-ready text. 12+ languages supported. Free, fast, and private — runs in your browser.
Drag and drop your files here
or Browse Files
Maximum file size: 10MB
Optical Character Recognition (OCR) is the technology that turns a picture of text into actual text — letters and words your computer can search, copy, and edit. Without OCR, a scanned PDF is just a sequence of images; you can read it, but you can't search it, copy passages out of it, or feed it into translation, summarization, or accessibility tools. Dokfo's free OCR tool fixes that. Upload any scanned PDF or photographed document, choose the language of the text, and Dokfo will analyze every page and produce a new PDF with an invisible searchable text layer overlaid on top of the original images. Visually the document looks identical, but now you can Ctrl+F to find any word, select and copy text from any page, and use the document with screen readers, translators, or AI tools. The OCR engine supports 12+ languages including English, Turkish, German, French, Spanish, Italian, Portuguese, Russian, Arabic, Chinese, Japanese, and Korean. It runs as WebAssembly inside your browser, so even confidential scans — IDs, contracts, medical reports — never leave your device.
How to OCR a PDF — Step by Step
1
Upload your scanned PDF or image file by clicking the upload area or dragging it onto the page. The file is loaded locally — no upload to any server.
2
Select the language of the text in your document. Choosing the right language dramatically improves accuracy, especially for non-Latin scripts like Arabic, Chinese, Japanese, or Korean.
3
Click "Extract Text". Dokfo runs the Tesseract OCR engine page by page in your browser, then produces a searchable PDF or extracted text file ready to download.
Why OCR PDFs With Dokfo?
12+ Language Support
Recognize text in English, Turkish, German, French, Spanish, Italian, Portuguese, Russian, Arabic, Chinese, Japanese, and Korean. Pick your language for the highest possible accuracy.
Searchable PDF Output
Get back a PDF that looks identical to your scan but is now fully searchable and selectable. The original images are preserved while a hidden text layer makes the content machine-readable.
Browser-Based Privacy
OCR runs entirely on your device using Tesseract compiled to WebAssembly. Confidential scans — IDs, contracts, medical records — never leave your computer.
No Sign-Up, No Watermark
Process scanned PDFs without creating an account. No watermark, no branding, no upload limit — just clean, searchable output.
When to OCR a PDF
- Make a scanned book, archive, or research paper searchable so you can find specific topics or quotes instantly.
- Extract text from old paper invoices, receipts, or contracts that were digitized as image-only PDFs.
- Prepare a scanned PDF for translation, summarization, or AI processing — these tools need real text, not images.
- Make scanned documents accessible to visually impaired users by giving them a real text layer that screen readers can speak.
- Convert photographed whiteboards, classroom notes, or handwritten signs into editable text you can search and reuse.
Your Scans Stay on Your Device
Dokfo's OCR tool uses Tesseract.js — the open-source Tesseract OCR engine compiled to WebAssembly — running inside your browser tab. The scanned PDF is read into memory, page images are extracted, and OCR runs locally on each page. The recognized text is then injected back into a new PDF, all without any server round-trip. This is essential for documents containing personal data, IDs, medical records, or confidential business information that you would never want to send to a third-party cloud OCR service.
Related PDF Tools
Frequently Asked Questions About OCR
What is OCR and how does it work?
OCR (Optical Character Recognition) is software that looks at images of text — like scanned pages or photos of documents — and identifies the letters and words inside them, producing real digital text you can search, copy, and edit. Dokfo uses Tesseract, the most widely deployed open-source OCR engine, running locally in your browser via WebAssembly.
Which languages does the OCR support?
Dokfo's OCR supports English, Turkish, German, French, Spanish, Italian, Portuguese, Russian, Arabic, Chinese (Simplified and Traditional), Japanese, and Korean. Choose the right language before processing — accuracy is significantly higher with the correct language model.
How accurate is OCR?
Accuracy depends on scan quality. Clear, high-resolution scans (300 DPI or higher) of printed text in standard fonts achieve 95-99% accuracy. Lower-resolution scans, photographs taken at angles, or unusual fonts may have more errors and require manual cleanup.
Does OCR work on handwritten text?
OCR is designed primarily for printed text. Handwritten text recognition is much less reliable — clear block printing may sometimes work, but cursive or messy handwriting will produce poor results.
Is the original PDF modified?
No. Dokfo's OCR produces a new PDF with a searchable text layer added on top of the original page images. The visual appearance is identical to the source, but text is now selectable and searchable.
Are my scanned files uploaded?
No. OCR runs in your browser using WebAssembly — Tesseract.js. Open Developer Tools → Network tab while processing to verify zero data transmission.
Can I OCR a PDF in multiple languages at once?
Yes — Tesseract supports loading multiple language models. If your document mixes languages (e.g., English and Turkish), you can select multiple language packs for combined recognition.
How long does OCR take?
Roughly 2-10 seconds per page on a modern laptop, depending on page complexity, resolution, and language. Long documents (100+ pages) may take a few minutes — keep the tab open until processing completes.