Try to search a scanned PDF and nothing comes up. Try to select text and you cannot — the cursor just draws a selection box like you are highlighting an area of a photo. That is because a scanned PDF is exactly that: a photograph of a page, stored inside a PDF container. To a computer, there is no text in the document — just pixels arranged to look like letters.
OCR (Optical Character Recognition) solves this. It analyses the image of each page, identifies the shapes of letters and words, and converts them into actual text that can be searched, copied, and edited. The technology has improved dramatically over the past decade and works well on most clearly printed documents.
How to make a scanned PDF searchable for free
The quickest way is to convert the scanned PDF to Word using Filero's free PDF to Word tool, which automatically runs OCR on scanned documents:
- Open the PDF to Word tool and upload your scanned PDF.
- The tool detects that the document is scanned and automatically activates OCR mode.
- The text is extracted from each page and placed into an editable Word document.
- Download the .docx file — you can now search, copy, and edit all the text.
No account needed. OCR runs entirely in your browser using Tesseract.js — your document is never uploaded to a server.
How OCR actually works
OCR software looks at each page as an image and tries to identify characters based on their visual shape. It works by breaking the image into regions, identifying likely letter boundaries, comparing shapes against a trained model of what each letter looks like, and assembling the results into words and sentences.
Modern OCR is trained on millions of document images and handles most standard typefaces accurately. Where it struggles is with handwriting (too variable in shape), very small or decorative fonts, low-quality scans (blurry or faded text), and documents that were not scanned straight (skewed pages throw off the character detection).
What kinds of documents OCR handles well
OCR works best on documents with clean, standard typefaces — printed contracts, typed letters, government forms, invoices, receipts, and reports. These are the kinds of documents that were originally produced on a computer, printed, and then scanned. The OCR engine has seen millions of examples of those fonts and recognises them reliably.
It also works reasonably well on older typed documents from typewriters, though accuracy depends on the condition of the original and the quality of the scan.
Handwritten documents are a different situation. Basic handwriting OCR exists, but accuracy is much lower because handwriting varies so much between individuals. A neatly printed form might be recognised well; a cursive note probably will not.
Getting better OCR results
- Scan at a higher resolution — 300 DPI is the standard for good OCR. Lower resolutions make characters harder to distinguish. If you are scanning a document yourself, use 300 DPI or higher.
- Keep the document flat and straight — a page that is rotated even a few degrees introduces errors. Make sure the document is straight on the scanner bed or table when photographing with a phone.
- Good lighting for phone scans— if you are photographing a document with your phone, use good even lighting without shadows. Scanner apps like Microsoft Lens or Apple's document scanner in the Files app apply automatic corrections that improve OCR results significantly.
- Check and correct the output — OCR is not perfect. Always review the extracted text, especially for names, numbers, dates, and technical terms. These are the areas where misrecognition is most likely.
Alternatives for making PDFs searchable
Adobe Acrobat Pro has built-in OCR under Tools > Enhance Scans. It creates a searchable PDF (with the text layer overlaid on the original image) rather than converting to Word. This preserves the exact visual appearance of the original scan while adding a searchable text layer — useful when you need the document to look exactly like the original but also be searchable.
Google Drive also offers basic OCR: upload a PDF to Drive, right-click and open with Google Docs, and Drive will attempt to extract the text. The results are useful for simple documents but often need significant cleanup for complex layouts.
For bulk processing of many scanned documents, Adobe Acrobat Pro and ABBYY FineReader are the professional tools of choice. For occasional personal use, the free browser-based approach via Filero is the fastest option.
Frequently asked questions
How accurate is free OCR?
For clean, high-resolution scans of standard printed documents, accuracy is typically 95 to 99 percent. That sounds high, but on a 10-page contract with 500 words per page, a 1 percent error rate means 50 mistakes — worth reviewing manually. For rough drafts or reference use, it is usually accurate enough without correction.
Can OCR read handwriting?
Standard OCR has limited accuracy with handwriting. Printed block letters are recognised reasonably well; joined cursive writing is much harder. Specialised handwriting recognition tools exist but are generally not free.
My document has both scanned and digital pages — will OCR work?
Yes. Filero's PDF to Word tool detects which pages are scanned and applies OCR selectively. Digital pages (pages with real text already) are converted directly without needing OCR.
Can I search the PDF without converting to Word?
For that, you need a tool that creates a searchable PDF (with an invisible text layer) rather than converting to Word. Adobe Acrobat Pro does this. The Filero tool converts to Word, which is useful when you want to edit the content — not just search it. If you only need to search and copy, the Word document can be converted back to PDF afterwards using Filero's Word to PDF tool.
Ready to try it?
Use Filero's free PDF to Word tool. No account needed, works on any device.
Open PDF to Word