OCR PDF: Make a Scanned PDF Searchable

Turn a scanned or image-only PDF into a searchable one. Text is recognized with the Tesseract engine in your browser and placed as an invisible layer behind each page image, so the scan looks the same but can be searched and copied.

Drag & drop a PDF here

Scanned documents and image-only PDFs work best

PDF

Overview

What is OCR PDF?

OCR PDF reads the text in a scanned PDF and saves a version you can search, select and copy from. The pages look exactly as they did; the difference is a hidden layer of real text sitting behind each scanned image.

A scanner or phone camera produces pictures of pages. To a computer they are just pixels, so Ctrl+F finds nothing and copying selects the whole image. Optical character recognition (OCR) looks at the shapes in the image and works out which letters they are. This tool uses Tesseract, an open-source OCR engine, compiled to WebAssembly and running in your browser.

How it works

How to make a scanned PDF searchable

  1. Open the scanned PDF

    The tool checks each page and tells you if some pages already have selectable text.

  2. Choose language and pages

    Select up to three languages used in the document and limit the run to the pages you need.

  3. Recognize text

    Each page is rendered at 200 or 300 dpi and passed to Tesseract. You can cancel between pages.

  4. Check and download

    Review the extracted text and confidence, then download the searchable PDF or a TXT file.

The text layer

How the searchable layer lines up with the scan

Tesseract returns every word with a bounding box measured in pixels. The tool converts each box to PDF points and writes the word there as invisible text (PDF text render mode 3). Font size comes from the box height, and each word is stretched horizontally to match the box width.

That is why selection highlights cover the right words on the image, and why search results jump to the correct place on the page. Pages stored with a rotation flag are handled, so text on a page shown in landscape still runs the right way.

Recognized word
Invoice
Box at 300 dpi
x 300–504, y 470–520 px
Written to the PDF
72 pt from left, 12 pt tall, invisible

Examples

Examples: what OCR does well and where it struggles

Source Typical result What helps
Office scan of a printed letter, 300 dpi Very few errors; confidence usually above 90% Default settings
Phone photo of a receipt Good on large text, mistakes on faded thermal print Flat, well-lit photo; "More accurate" setting
Two-column journal article Text found; reading order in TXT may jump between columns Check the TXT export before reusing it
Handwritten notes Mostly unreadable Tesseract is trained on printed type, not handwriting
Tables and forms Words found, but table structure is not kept Use PDF to Excel on files that already have text

Use cases

Use cases for OCR on PDFs

Finding things in scanned paperwork

Search a 60-page scanned lease for "deposit" or a policy number instead of reading every page.

Ctrl+F "deposit" 3 matches

Quoting from old documents

Copy a paragraph from a scanned report or a book chapter and paste it into your notes, then correct any misread characters.

scan.pdf scan-ocr.txt

Archives and document systems

Desktop search, Windows indexing and most document management systems can index searchable PDFs but not image-only ones.

image-only indexed

Tips

Tips for better OCR accuracy

  • Select only the languages that appear in the document. Extra languages slow OCR down and can increase mistakes.
  • Use "More accurate (300 dpi)" for small print, footnotes and faint scans.
  • Rotate sideways pages first with Rotate PDF so the text reads upright.
  • Skip pages that already have selectable text to avoid duplicate search results.
  • Treat the output as a draft: check names, numbers and amounts before relying on them.

Limitations

Accuracy limits and other limitations

OCR is never perfect

Similar characters such as 0/O, 1/l and rn/m get confused, especially on low-quality scans. The confidence score is shown so you can judge the result.

Languages

English, Spanish, French, German, Portuguese, Italian, Dutch, Turkish and Hindi are offered. Hindi text is recognized and included in the TXT export, but it is not added to the PDF's searchable layer, because the embedded font covers Latin, Greek and Cyrillic scripts only.

Speed and memory

Expect a few seconds to around 15 seconds per page depending on the device. Long documents are best processed in page ranges on phones.

Layout is not rebuilt

The tool adds searchable text; it does not turn the scan into an editable Word document or keep table structure.

Privacy

Privacy and downloads

Your PDF is rendered and recognized in this browser tab and is not uploaded to TryFormatter. The first time you run OCR, the browser downloads the Tesseract engine and the language model files from the public jsDelivr CDN; those downloads contain no part of your document. Browsers usually cache them for the next run.

Related tools

Related tools

Frequently Asked Questions

Will OCR change how my scanned pages look?

No. The scanned images are kept exactly as they are. The recognized text is added as an invisible layer behind them, so the pages look the same but can be searched and selected.

How accurate is the OCR?

On clean 300 dpi scans of printed text, Tesseract usually gets nearly every word right. Accuracy drops on blurry photos, faded print, unusual fonts and handwriting. The result panel shows an average confidence score; below about 70% you should expect noticeable errors.

Why do I see duplicate search results after OCR?

The page probably already had a text layer. OCR adds a second one, so each word is found twice. The tool warns you when it detects existing text; limit the page range to the pages that are actually scanned images.

Can I OCR a PDF in a language that is not listed?

Not yet. Only the listed languages are offered, because they have tested language data packages. Picking the wrong language usually gives poor results, so choose the closest match only if the alphabet is the same.

Does OCR work offline?

After the engine and the language files have been downloaded once, the browser normally caches them. The first run needs an internet connection to fetch those files; your PDF is never part of that download.

Related Tools

PDF to Word

Convert PDF documents to editable Microsoft Word (.docx) files online for free. No upload, and the conversion runs entirely in your browser.

PDF Editor

Add text, images, shapes, drawings and highlights to PDF pages, edit them with undo and redo, and download the result.

PDF Compressor

Reduce PDF file size using simple quality presets and optional advanced controls. Balance document clarity and size reduction directly in your browser.

Organize PDF

Reorder, rotate, duplicate and delete PDF pages with a visual page grid, insert pages from another PDF, and download the result.

PDF to Image

Convert PDF pages to JPEG or PNG images online for free. No upload needed, runs in your browser, download as ZIP.

Image to PDF

Convert images to PDF online for free. Combine JPG and PNG files into one PDF document. No upload, runs in your browser.

Compress PDF to Exact Size | 100KB, 200KB, 500KB

Compress a PDF to 100KB, 200KB, 500KB, 1MB, 2MB, or a custom limit. Fit under portal upload caps with automatic quality optimization.

PDF Merger

Merge multiple PDF files into one document for free. Runs entirely in your browser with no server upload.

Expert Guides & Tutorials

The End of Cloud-Based PDF Editors: Security Risks of Server-Side Document Processing

The End of Cloud-Based PDF Editors: Security Risks of Server-Side Document Processing

Cloud-based PDF editors expose your sensitive business documents to unnecessary risks. Learn how browser-native WebAssembly tools are replacing legacy server-side PDF processing in 2026.

Read Guide →