Free and instant — plug in your own numbers.
A scanned PDF is a picture of a page. You can read it, but your computer can't: a search finds nothing, and you can't select or copy a single word. OCR (optical character recognition) fixes that. It reads the letters in the picture and adds them to the file as hidden text, so the pages look exactly the same but you can now search, highlight and copy them.
Calcnio's free OCR PDF tool does this in your browser. The file is read on your own device and never uploaded, and it accepts photos of pages (JPG, PNG or WebP) as well as PDFs. On a Mac, Preview can do the same job, and Google Drive can turn a scan into an editable Google Doc.
How to tell if your PDF needs OCR
Open the PDF and press Ctrl+F (Cmd+F on a Mac). Search for a word you can see on the page. If the search finds nothing, or you can't drag to select a line of text, the page is an image with no text inside, and it needs OCR.
Scanners, phone scanning apps and PDFs made from photos usually produce this kind of file.
It matters for more than search. Screen readers read text, not pictures, so a scan with no text layer can't be read aloud to someone who is blind or has low vision. The federal Section 508 program's training on scanned documents says OCR has to be done before a scanned PDF can be made accessible.
How to make a scanned PDF searchable with Calcnio
- Open the OCR PDF tool and drop your scanned PDF on the box, or press Choose files. You can add photos of pages too.
- Pick the language of the text. English is ready at once. Spanish, French, German and Portuguese download the first time you pick them, about 1 MB each.
- Decide what to do with pages that already have text. By default they are left as they are, and only scanned pages are read. If a page mixes typed text with a scanned part, choose Read them too.
- Press Make searchable. A bar shows which page is being read, and you can press Cancel at any time.
- Download the new PDF, open it and press Ctrl+F to test a search.
The first run downloads the text reader once (about 5 MB) from Calcnio, and your browser keeps it for next time. In Calcnio's own test, a 10-page scan at 300 DPI took about 10 seconds on a computer. A phone takes longer, as it reads one page at a time.
What OCR changes in your file, and what it doesn't
The scan stays the same picture at the same quality. The tool adds the words it reads as invisible text, placed over the matching words on the page. That is why the PDF looks exactly as before but now lets you search, highlight and copy. Most scanners that make "searchable PDFs" work the same way.
After the run, the tool shows how many words it found and an average confidence score. Open Show the text that was found to check the text before you rely on it. If a page scores below 60% confidence, the tool names that page, so you know where to look.
Key idea OCR makes a best guess at every word. On a clean, straight scan it is usually right, but check names, account numbers and amounts against the page before you copy them anywhere.
Scan quality: the numbers that matter
OCR reads pixels, so the resolution of your scan, measured in dots per inch (DPI), decides how many pixels each letter gets. Here is the math for a US letter page, which is 8.5 × 11 inches. Type is measured in points, and a point is 1/72 of an inch, so 10-point type is 10 ÷ 72 = 0.139 inches tall.
| Scan resolution | Page size in pixels | 10-point type, in pixels | 8-point fine print, in pixels |
|---|---|---|---|
| 72 DPI | 612 × 792 | 10 | 8 |
| 150 DPI | 1,275 × 1,650 | 21 | 17 |
| 200 DPI | 1,700 × 2,200 | 28 | 22 |
| 300 DPI | 2,550 × 3,300 | 42 | 33 |
| 600 DPI | 5,100 × 6,600 | 83 | 67 |
The pixel figures are for the font size; the letters themselves are a little smaller. Google's help page for Drive OCR asks for text at least 10 pixels high, and at 72 DPI ordinary 10-point type only just reaches that. At 300 DPI the same type gets about 42 pixels. Calcnio's tool read every word correctly on a clean, straight 300 DPI scan in its tests, and scans under 200 DPI give more mistakes.
Going higher has a cost. A 600 DPI scan has twice as many pixels across and twice as many down as a 300 DPI scan, so four times as many in total, and the file grows with them. For printed text, 300 DPI is the usual choice. If the PDF ends up too big to send, see how to compress a PDF for email.
Other things that hurt OCR:
- Blur, shadows and creases. A sharp, evenly lit page reads best. Calcnio's Scan to PDF tool can clean up phone photos first.
- Sideways or upside-down pages. The text must be the right way up. Turn pages with Rotate PDF before you run OCR.
- Handwriting and unusual fonts. Printed text works far better than handwriting.
- The wrong language. Pick the main language of the document.
Other free ways to OCR a PDF
Preview on a Mac
Apple's guide to embedding text in a PDF gives the steps: open the PDF in Preview, choose File > Export, then select Embed Text. Preview finds the text and embeds it so it can be selected in any PDF viewer. The option only appears if no text was recognized in the document before.
Google Drive
Google's guide to converting PDF and photo files to text says to open Google Drive on a computer, right-click the file and choose Open with > Google Docs. You get a new Google Doc, not a searchable PDF. Google's tips: the file should be 2 MB or smaller, text at least 10 pixels high, the page right side up, and common fonts such as Arial or Times New Roman work best. Bold, italics, font size and line breaks are likely kept, but lists, tables, columns and footnotes are not likely to be detected. Because the file sits in your Drive, it is stored on Google's servers.
Microsoft Word on Windows
Microsoft's guide to inserting scanned text into Word says to choose File > Open and pick the PDF. Word tells you it will try to import the text, and you get an editable document. The guide covers desktop Word on Windows, notes that handwriting is seldom recognized, and tells you to always proofread the result. For scanned tables, our guide on converting a PDF to Excel shows how to get them into a spreadsheet instead.
OneNote
In OneNote, right-click a picture and choose Copy Text from Picture, then paste the text where you want it. For a printout with several pages, choose Copy Text from this Page of the Printout or Copy Text from All the Pages of the Printout. Microsoft's OneNote OCR guide notes that the command may not appear at once for complex images. This copies text out; it doesn't make your PDF searchable.
| Method | Works on | What you get | Pages look the same? |
|---|---|---|---|
| Calcnio OCR PDF | Any modern browser | Searchable PDF | Yes |
| Preview | Mac | Searchable PDF | Yes |
| Google Drive | Any browser, with a Google account | Editable Google Doc | No |
| Word | Windows, desktop Word | Editable Word document | Formatting may change |
| OneNote | Windows and Mac | Copied text | Not applicable |
Get the text out of the PDF
Once the PDF is searchable, you may want the words themselves, for an email, a spreadsheet or a notes app. Run the OCR'd file through the PDF to Text tool. It saves the text as a .txt or Markdown file, joins lines into paragraphs, keeps each table row on one line with tabs between the cells, and marks where each page starts. If you skip OCR, PDF to Text tells you which pages have no text to copy.
When OCR is worth the effort
- Tax and money records. The IRS says you generally need to keep the records that support your tax return for 3 years, and longer in some cases; see its page on how long to keep records. A searchable PDF of a year's receipts lets you find one store name or invoice number in seconds.
- Contracts and leases. Search for "deposit" or "notice" instead of scrolling through 30 pages.
- Files you share with others. OCR is the first step to making a scan readable with a screen reader. Section 508 guidance adds that a fully accessible PDF also needs tags and alternative text for images, which OCR alone doesn't add.
Tax, medical and bank papers are exactly the files you shouldn't hand to a site you don't know. To keep a year of scanned receipts in a few tidy files, see how to merge scanned receipts for tax time. If your documents are still phone photos, start with our guide on how to convert JPG to PDF.
Frequently asked questions
Can OCR read handwriting?
Not well. OCR is built for printed text, and Microsoft's own guidance says handwriting is seldom recognized. Expect to type handwritten notes yourself.
Is OCR text always accurate?
No. Accuracy depends on the scan. A sharp, straight page at 300 DPI reads far better than a dark or blurry phone photo, so check the confidence score and the key numbers.
Can I OCR a PDF on my phone?
Yes. Calcnio's OCR PDF tool runs in a phone browser too. It takes longer than on a computer, because it reads one page at a time.
A quick checklist
- Press Ctrl+F and search for a word you can see. No match means the page needs OCR.
- Scan at 300 DPI, straight and in good light.
- Turn any sideways pages the right way up.
- Run the file through the OCR PDF tool with the right language.
- Check low-confidence pages, names and numbers against the scan.
- Need the words on their own? Save them with PDF to Text.
This article is for general education and is not financial advice. Rules, rates and products vary by lender and country. See our disclaimer.