PDF to Word Converter

Turn a text-based PDF into an editable .docx with headings, lists, tables and pictures. The file is read on your device and never uploaded. Fidelity is lower than paid desktop tools, so check the preview first.

PDF file

No PDF yet

Pick a PDF and see the editable text before you download the Word file.

Layout
Lines are joined into paragraphs that re-wrap. Best for editing.
What to rebuild

Word file

Pages
—
Words
—
Tables · Images
—
Add a PDF to start

The PDF is read and the .docx is written on your device. Nothing is uploaded.

Layout fidelity is lower than paid server tools: floating text boxes, charts, merged cells and unusual layouts are not reproduced exactly. Works best on text-based PDFs. Check the preview before you download.

How to convert a PDF to Word

This PDF to Word converter reads your PDF in the browser and writes a .docx you can edit in Word, LibreOffice or Google Docs. The file is not sent anywhere. Four steps:

  1. Choose the PDF or drop it on the box. Use Try a sample PDF if you only want to see what the output looks like.
  2. Pick Flowing text to edit the content, or Keep page layout to stay close to the printed page. Switch off anything you do not want rebuilt, such as images.
  3. Read the preview. It shows the text, headings, lists and tables the way they will land in the Word file. If the text is garbled here, it will be garbled in Word too.
  4. Press Download .docx. The file is named after the PDF, so report.pdf becomes report.docx. Copy text gives you the plain text if that is all you need.

To convert only part of a file, type pages such as 1-3, 5, 8-. An empty box means every page.

Why does a PDF lose its formatting in Word?

A PDF is a drawing instruction list: put this word at this spot, in this font. It does not know what a paragraph, a heading or a table is. Every converter, including paid ones, has to infer those from positions. Lines that start at the same left edge and end near the right margin are probably one paragraph. A line set in a bigger font is probably a heading. Text that lines up in columns across several rows is probably a table.

Inference works well on ordinary documents: reports, letters, contracts, articles, statements. It works badly on brochures, posters and slides, where text floats around pictures and nothing follows a grid. Expect to spend a minute fixing those, and for those files a paid desktop converter will usually do better than any browser tool, this one included.

Flowing text or keep page layout?

Flowing textKeep page layout
LinesJoined into paragraphs, hyphens at line ends removedOriginal line breaks kept
Page size and marginsTaken from page 1, margins evened outEach page keeps its own size and margins
ColumnsRead in order, left column first, as one columnReal Word columns
Page breaksOptionalOne Word page per PDF page
EditingText re-wraps normallyFixed lines, so adding words shifts layout

If you plan to rewrite or reuse the text, choose flowing. If you need the same page count and look, for example to fill in or annotate a form, choose keep layout and accept that editing is clumsier. In our test with a five-page Indonesian report, keep layout came back as five pages in LibreOffice, matching the PDF. Word may break a line or two differently if it lacks the original font.

What is rebuilt, and what is not

In the PDFIn the Word file
Larger text on its own lineHeading 1, 2 or 3, the largest size being level 1
Bold and italic fontsBold and italic runs, found from the font name
Bullets and 1. 2. 3. listsReal Word lists with nesting
Rows that line up in columnsA Word table, header row marked, numbers right-aligned. Borders are always added
PicturesPlaced in reading order. Photos are saved as JPEG, transparent ones as PNG, and large ones are scaled to 2000 px
Repeated headers, footers, page numbersSkipped by default, can be kept
Text boxes, shapes, charts drawn as linesNot reproduced
Text colours, links, form fieldsNot carried over
Text printed sidewaysSkipped, with a note

A full-page picture behind real text, which is how many searchable scans are built, is left out so the page is not shown twice.

How do I convert a scanned PDF to Word?

A scan is a photo of a page. There are no letters to extract, so the preview is empty and the tool says the PDF looks scanned. Press Read pages with OCR, pick the language of the document and wait. The OCR engine (Tesseract) and the language data are downloaded the first time, then the pages are recognised on your device. A clean 300 DPI scan of a printed page is usually readable with a few errors. A crooked phone photo, handwriting or a table full of small numbers is not, so proofread the result and check every figure against the original.

Large files, page ranges and other scripts

Pages are read one at a time with a progress bar and a Cancel button. A 120-page, 26 MB test file took about four seconds to read on a desktop browser, and the Word file came out at 5.5 MB because the pictures are recompressed. PDFs up to 150 MB and 600 pages per conversion are accepted. For longer documents use page ranges.

Indonesian accents, Spanish characters, Chinese and Arabic come through as normal Unicode text, and Arabic paragraphs are marked right-to-left so they align correctly. The Word file asks for the same font family as the PDF. If your computer does not have it, Word substitutes another, so line breaks and letter shapes can differ a little. A PDF with a broken font table can produce nonsense characters, and the preview shows that before you download.

Frequently asked questions

How do I convert a PDF to Word without uploading it?

Drop the PDF on this page, check the preview and press Download .docx. Your browser reads the PDF and writes the Word file itself, so the document never leaves your phone or computer. A 10-page text PDF usually takes a few seconds.

Why does my PDF lose its formatting when I convert it to Word?

A PDF stores where each piece of text sits on the page, not paragraphs, so any converter has to guess the structure. This one rebuilds lines into paragraphs, headings from font size, lists and simple tables. Floating text boxes, charts drawn as vector shapes and unusual layouts will not come out exactly as in the PDF.

Can I convert a scanned PDF to Word?

Yes, with OCR. A scan is a picture with no text in it, so the tool detects that and offers to read the pages with Tesseract OCR in your browser. Pick the language first. Expect to proofread: OCR on a clean 300 DPI scan is usually good, a skewed phone photo much less so, and tables from scans are rebuilt less reliably.

How do I convert only some pages of a PDF to Word?

Type the pages in the page box, for example 1-3, 5, 8- (8- means page 8 to the end). Leave it empty for the whole file. Only those pages are read and exported, which also makes big PDFs faster.

Does the Word file keep tables and images?

Tables are detected when rows line up in columns, and they come out as real Word tables with borders added. Pictures embedded in the PDF are extracted and placed in reading order. A full-page background image behind real text is skipped so the page is not duplicated. Charts that are vector drawings are not exported as pictures.

What is the difference between flowing text and keep page layout?

Flowing text joins lines into normal paragraphs that re-wrap when you edit, and removes hyphenation at line ends. It is the better choice for editing. Keep page layout copies the page size, margins, line breaks, indents and real Word columns so each page looks closer to the PDF, but lines stay fixed and editing is less comfortable.

Is there a file size or page limit?

You can open PDFs up to 150 MB and convert up to 600 pages at once. Pages are read one at a time with a progress bar and a Cancel button, so a 100-page file does not freeze the tab. For longer documents convert them in page ranges.

Will Arabic, Chinese or Indonesian text convert correctly?

Text with a proper text layer is carried over as Unicode, including Indonesian accents, Arabic (marked right-to-left) and Chinese. The fonts in the Word file are named after the PDF fonts, and Word substitutes ones you do not have, so letter shapes can differ a little. PDFs that use damaged font encodings may give garbled text, which the preview shows before you download.