Scanned PDF to Word (OCR)

Convert a scanned PDF to an editable Word document in your browser. Every page is read with OCR and rebuilt as paragraphs and tables. Free, no upload.

Runs in your browser. Your files are never uploaded. How it works

This page converts a scanned PDF into an editable Word document. Every page is read with text recognition (OCR), and the text is rebuilt as paragraphs, headings and tables you can edit in Word, Google Docs or LibreOffice. The recognition runs in your browser, so the scan never leaves your device.

Convert a scanned PDF to Word

  1. Click Choose a PDF and pick the scan. Type the password if it asks for one.
  2. Pick the Text language if it is not covered by Auto.
  3. Click Convert to Word. The first time, the text recognition engine (about 11 MB) and, with tables on, the table engine (about 19 MB) are downloaded and cached.

Scans that convert well

Document Result
Printed letters and reports Editable paragraphs
Old contracts or forms on paper Text you can update and reuse
Printed tables and price lists Word tables with rows and columns
Photos of pages saved as PDF Editable text, if the photo is straight and sharp

Tips for better results

  • Scan at 300 DPI if you can, straight and in good light.
  • Turn sideways pages the right way up first with Rotate PDF.
  • Only need the text, or a searchable PDF? Use OCR PDF.
  • Proofread the result: text recognition can confuse similar letters and numbers, such as 0 and O.

Frequently asked questions

Why is every page read as a scan on this page?

Read every page as a scan is selected here, so each page is read from its image, even if the PDF has a broken, partial or garbled text layer. For a normal PDF made by software, choose Use the PDF's text instead; it is faster and exact.

Which languages can it read?

Auto covers English, European languages, Chinese and Japanese. For Korean, Russian, Arabic, Hindi, Thai, Greek, Tamil or Telugu, pick the language under Text language before converting.

Is bold and italic kept from a scan?

No. Text recognition reads the words, their size and their position, not the font style, so scanned text comes in as plain text. Paragraphs, headings (from larger text) and tables are still rebuilt.

How accurate is it?

Clean, straight scans of printed text read well. Small print, low-resolution scans, handwriting, stamps and photos taken at an angle give more mistakes. Proofread names, numbers and amounts.

Is my scan uploaded?

No. Text recognition runs in your browser, on your device. The recognition models are downloaded once and cached; your document is never sent anywhere.

Related tools

Excel to PDFTurn an Excel spreadsheet (.xlsx) into a PDF with selectable text, right in your browser.PDF to PowerPointTurn each PDF page into a PowerPoint slide with editable text, right in your browser.Word to PDFTurn a Word document (.docx) into a PDF, right in your browser. Nothing is uploaded.PDF to ExcelFind the tables in a PDF, scan or image and download them as Excel or CSV. Runs in your browser.JPG to PDFCombine JPG, PNG and other images into a single PDF, in the order you choose.PDF to JPGTurn every page of a PDF into a JPG or PNG image, one page or all at once.