OCR: make a scanned PDF searchable
Drop a scanned PDF or an image, pick the language (English, French or both) and start the OCR. The recognized text is placed invisibly behind each page, so you can search, select and copy it while the document looks exactly the same. The work happens in your browser.
- ✅ Free
- 🔒 No sign-up
- 📱 Works on mobile
- ⚡ Instant results
🔒 Everything runs in your browser. Your data and files never leave your device.
📌 Key facts
- OCR PDF adds an invisible text layer to a scanned document without changing the page image.
- The languages offered are English, French or both together.
- At 200 dpi, the recommended resolution, expect roughly 6 to 12 seconds per page.
- Sharp printed text often reaches over 95% correct characters; handwriting still reads poorly.
- Recognition runs in your browser, so the document is not uploaded to a server.
What OCR does to a PDF
OCR (optical character recognition) reads the image of a page and finds the letters. A scanned PDF is just a picture: you cannot search for a word or copy a paragraph. After OCR, an invisible text layer sits over each word at its exact position.
- Drop the PDF or image and choose the language. "French + English" suits mixed text.
- Start the OCR: a progress bar follows each page.
- Download the searchable PDF, or copy the plain text.
Pages that already hold text are skipped by default, which avoids redoing work and keeps the file small.
Settings and expected quality
| Resolution | Use | Typical time per page |
|---|---|---|
| 150 dpi | Clean documents, large text | 3 to 6 seconds |
| 200 dpi | Letters, receipts (recommended) | 6 to 12 seconds |
| 300 dpi | Small print, average scans | 12 to 25 seconds |
Sharp printed text often reaches over 95% correct characters. Handwriting, skewed photos and busy backgrounds read poorly: straighten the image first with the Scan to PDF tool.
What you get
The PDF keeps the exact look of the scan: the image is neither recompressed nor changed. Only an invisible text layer is added. Characters outside the standard Latin set are left out of that layer. The plain text is also available for download so you can paste it into a word processor.
OCR PDF: which documents it works best on
OCR PDF helps whenever a document is only a picture of text. For example, a scanned contract, an old class handout or a letter that arrived as an image. Then you can search a word with Ctrl+F and copy a passage.
- Scanned contract or invoice: find an amount or a name in seconds.
- Photographed textbook page: copy an excerpt into your word processor.
- Paper archive: a searchable folder beats a stack of sheets.
How to prepare a scan before running OCR
Scan quality decides text quality. First, straighten the page and avoid shadows. Then pick the right language so accents and spelling match.
| Problem | Effect on OCR | Fix |
|---|---|---|
| Crooked page | Cut words or sideways lines | Straighten the image first |
| Gray or colored background | Errors on thin letters | Rescan in black and white |
| Low resolution | Letters mixed up | Move up to 300 dpi |
| Mixed languages | Missing accents | Choose English + French |
Finally, proofread key numbers, because a 5 can pass for a 6. For an amount or an account number, always check the original.
What to do after OCR PDF
Once the PDF is searchable, you can do more with it. If the photo has a poor angle, start with scan to PDF. To cut the file size, run compress PDF next. Want to edit the text in Word? Use PDF to Word.
To check two versions of the same contract, open compare PDF. Last, before you share the file, hide private data with redact PDF.
OCR PDF: common mistakes and quick checks
After an OCR PDF run, always do a simple test. First, open the file and search for a word you can see on the page. If the search finds it, the text layer works. Otherwise, run the OCR PDF again at a higher resolution.
| Symptom | Cause | What to do |
|---|---|---|
| Search finds nothing | Wrong language chosen | Run again with the right language |
| Words stuck together | Scan too pale | Rescan with more contrast |
| Page skipped | It already has text | That is normal, nothing to do |
| Slow run | 300 dpi on many pages | Drop to 200 dpi |
Next, copy one paragraph into your word processor. You will see quickly whether the accents come out right. Finally, keep the original scan, which stays your visual proof. OCR PDF only adds a text layer, so nothing gets lost.
Do you work on a document with dozens of pages? Then test three pages first. That way you set the language and the resolution without waiting for the whole file.
When OCR PDF is the wrong tool
OCR PDF cannot read everything. For example, cursive handwriting, tiny stamps and faded fax pages give weak results. In those cases, retyping the key lines is faster. Also, a PDF that already has real text does not need OCR, so the tool skips those pages on purpose.
For example, for a one-page letter, 200 dpi is enough. You start the OCR, wait a few seconds, then download the PDF. Next, you search a word and copy the text. After that, you file the document. So you find each document without opening ten pages. Then OCR PDF saves you time every week. Finally, you keep the original untouched. Also, you glance at the key numbers. Then you move the file to the right folder. If you share it, your colleagues can search it too. They save time as well.
3 real-life examples
Find a word in a scanned contract
Your 12-page contract is a stack of images. After OCR in English, Ctrl+F finds "termination" in a moment and highlights it.
Copy text from a photo
You drop a photo of a book page, run OCR in English and click "Copy text" to paste the paragraph into your document.
A mixed file
Of 20 pages, 6 already hold text: they are skipped and the other 14 are read. The summary says "14 page(s) read, 6 skipped".