Skip to content
PageSaver

How to Make a Scanned PDF Searchable (Free OCR Options)

A scanned PDF is a stack of page pictures, so there's no text in it to search. To make it searchable, you run OCR (optical character recognition), which reads the letters in each picture and adds an invisible text layer on top. The page looks the same, but you can now select, copy and Ctrl+F the words. You can do it for free on a Mac with Preview, in Google Drive, on the command line with OCRmyPDF, or in your browser with our Scanned PDF to Text tool, which doesn't upload the file.

We make PageSaver, the browser extension, and PixelPDF, the free PDF tools on pagesaver.ai. This guide shows how to check whether your PDF really is scanned, how to add the text layer step by step, and what you give up when you do.

How to tell if a PDF is scanned

Open the PDF and try two things:

  1. Drag across a sentence. If you can't highlight individual words, or the whole page gets selected as one box, the page is an image.
  2. Press Ctrl+F (⌘+F on a Mac) and search for a word you can clearly see on the page. No match usually means there's no text layer.

For a more exact answer, open the file in PixelPDF Analyze PDF. The Text layer section labels the document as Text-based, Scanned, Image-based or Mixed, and lists the Pages needing OCR. That's useful for mixed files, like a typed report with a few signed pages scanned in at the end. The check runs in your browser and doesn't change the file.

What OCR does to a PDF

OCR looks at the image of each page, works out which shapes are letters, and records what it read along with where each line sits. A searchable PDF keeps the page image you see and places that text, invisibly, at the same position. When you search or select, you're working with the hidden text; when you look at the page, you see the original scan.

OCR is a guess, though usually a good one. Clean, straight, high-contrast scans read well. Faint copies, handwriting and tiny print are where mistakes show up, so check names and numbers against the page before you rely on them.

Make a scanned PDF searchable in your browser

Scanned PDF to Text is free and needs no account. Here's how it works:

  1. Open pagesaver.ai/tools/ocr and drop your PDF on the preview area, or click it to choose a file.
  2. Wait for the pages to appear in the preview.
  3. Pick a Recognition language: English, Simplified Chinese, Traditional Chinese, English + Simplified Chinese or English + Traditional Chinese. Use a mixed option when the same PDF has both English and Chinese.
  4. Click Run OCR. The status line shows each page as it goes.
  5. When it finishes, a summary tells you how many pages were processed, how many already had text, how many went through OCR, and how many words it found. Click Download searchable PDF.

The finished copy also shows in the preview area, so you can look it over before you download it. Once it's saved, open it and press Ctrl+F to confirm the text is there.

Pages that already have real text aren't OCRed again. The tool reuses their existing text, so a mixed PDF doesn't pick up recognition errors on pages that were fine to begin with.

What changes in the searchable copy

The result is a new PDF, not your original with text added. Each page is redrawn as an image with the text layer on top. That means some things don't carry over: forms, links, annotations, attachments, accessibility tags and digital signatures are not kept. Hold on to your original, and don't send the OCR copy of a signed contract as if it were the signed file.

Because every page becomes an image, the copy can be large. If it's too big to email, here's why PDFs get big and how to shrink them.

Limits worth knowing

  • Languages: English, Simplified Chinese and Traditional Chinese only. For a scan in Spanish, Japanese or another language, use one of the other options below.
  • Speed: recognition runs on your own computer, so a long scan can take a while.

Does the file get uploaded?

No. All processing happens in your browser, and the PDF isn't sent to our servers. The page still loads its code over the internet, but the document itself stays on your device. That matters for scans of IDs, bank statements and medical letters, which are exactly the documents people tend to scan. We wrote more about this in where your file actually goes.

Need Word or Markdown instead of a PDF?

If you want to edit the text rather than search it, skip the searchable PDF. PDF to Word and PDF to Markdown both have a Scanned pages option to read image-only pages with OCR, which is turned on by default. It only runs on pages without a usable text layer and offers the same language choices. Review names, numbers and tables in the result, since that's where OCR errors are easiest to miss.

Other ways to OCR a PDF for free

Any of these will do the job. Which one is best depends on your computer, your language and whether you mind the file leaving your device.

  • Preview on a Mac. Open the PDF, choose File > Export and select Embed Text. Apple says the text can then be selected in any PDF viewer, and that the option only appears if no text was previously recognized in the document. Preview's Live Text is a different feature: Apple documents it for copying text out of photos and images, it doesn't add anything to the PDF, and it isn't available in all regions or languages.
  • Google Drive. Upload the PDF, right-click it, and choose Open with > Google Docs. According to Google's help page, the file should be 2 MB or smaller, and lists, tables, columns, footnotes and endnotes are not likely to be detected. You get an editable Google Doc rather than a searchable PDF, and Drive detects the language for you.
  • Adobe Acrobat. The desktop app is paid. Open the scan, use the Scan & OCR tool to recognize the text, and choose the pages and language; menu names vary between Acrobat versions, so Adobe's help has the current steps. Adobe also has a free online OCR tool; your file is handled by Adobe's servers, and you need to sign in to download the result.
  • OCRmyPDF. If you're comfortable in a terminal, OCRmyPDF is open source (MPL-2.0), runs on your own machine using the Tesseract engine, and supports more than 100 languages through language packs. One command does it: ocrmypdf input_scanned.pdf output_searchable.pdf. It's the best choice for whole folders of scans.

When the built-in option is good enough

If you're on a Mac and Preview offers Embed Text, try that first. There's nothing to install and nothing to upload. If all you want is the words, not a PDF, and the file is small and not private, Google Docs works fine. Reach for a separate tool when Preview isn't available, when the scan mixes English and Chinese, or when you'd rather the document didn't go to anyone's servers.

Why can't I find a word in a PDF I just OCRed?

Usually OCR misread that word, so the hidden text says something slightly different from the page. Try searching for part of the word, or a nearby word that's printed more clearly. If a lot of searches fail, check that you picked the right recognition language, and if possible scan the page again at a higher quality. Faint photocopies and crooked phone photos are the most common culprits.

It's also worth checking that you're searching the new file. OCR tools like ours give you a new copy, and it's easy to open the original again by mistake.

Which one should you use?

For an English or Chinese scan you'd rather keep on your own device, use Scanned PDF to Text: drop the file in, pick the language, click Run OCR, and download the searchable copy. On a Mac, Preview's Embed Text is the quickest built-in route. For other languages or big batches, OCRmyPDF is hard to beat, and Acrobat makes sense if you already use it. Whichever you choose, keep the original and spot-check the text before you rely on it.