PDF to Word, ready to edit

What PDF to Word does

PDF to Word turns a PDF into a Word file, a .docx, that you can edit. Change a few words in a letter, update a CV you only have as a PDF, fix a date in a form, or reuse a report as the start of a new one. The text comes across as real text, in paragraphs that re-wrap as you type, with its headings, lists and pictures where they were.

Everything happens in your web browser. The page reads each page of your PDF with pdf.js, the PDF reader built into Firefox, works out which lines belong together, and writes the Word file itself, on your own phone or computer. Your PDF is never uploaded.

• Paragraphs, not lines: lines that simply ran on are joined into one paragraph, so the text flows again when you edit it. A word broken across two lines with a hyphen is joined back up.
• Headings, lists and columns: big or bold titles become Word headings, bullets become Word lists, and pages set in two columns get Word columns.
• Pictures in place: photos, charts and drawings go into the Word file as pictures, where they were on the page.
• Honest with scans: a scanned page has no words in it to copy, so it goes in as a picture, and the result tells you which pages those are.

Step by step

Here is the whole way, on a phone or a computer:

1. Add your PDF. Tap Choose PDF files, or drop PDFs onto the page. Each PDF gets a row with a picture of its first page, its number of pages and its size. You can add as many as you like.
2. Unlock it, if it asks. A PDF locked with a password asks for it on its own row. Type it and press Open.
3. Press Convert to Word. With several PDFs the button says how many, like Convert 3 PDFs to Word. While it works you can see how far it has got, like Page 12 of 117, and you can press Stop at any time: you still get a Word file with the pages read so far.
4. Download. Each PDF gives its own Word file, named after it: report.pdf becomes report.docx. With several PDFs there is also a Download all (ZIP) button, for a ZIP file with every Word file inside.
5. Open and edit. Open the file in Microsoft Word, Google Docs, WPS Office, LibreOffice or any other app that opens .docx files.

The result shows each Word file's size, its number of pages and how long it took. If some pages are scans, or anything else needs a look, a short note under the file says which pages and why. If you add or remove a PDF afterwards, the result is cleared and the Convert button comes back.

What comes across

A PDF only stores where each word and picture sits on the page. The converter looks at the page the way you would, and rebuilds the parts a Word file is made of:

Each page of the PDF starts a new page in Word, with its own margins and the spaces between paragraphs as in the PDF. That keeps the Word pages close to the PDF's pages: our 14-page research paper, set in two columns, still had 14 pages when we opened its Word file in LibreOffice.

Lines that end early on purpose keep their line breaks, like an address, a poem or the lines of a signature block. Lines of computer code keep each line and its indents, in a typewriter font.

What doesn't come across exactly

Some things in a PDF can't be rebuilt exactly, because the PDF never stored them as such. Here is what to expect, so nothing surprises you:

• Tables become lines with tab stops. Each row of a table becomes one line, with a tab stop where each column starts, so the columns line up and the text reads row by row. The table's lines and shading are left out, and a cell with several lines of text spreads over several lines. To turn the rows into a real Word table, see Editing the Word file.
• Fonts are swapped for common ones. A PDF's fonts are not copied into the Word file. Text keeps its font when it is one that comes with Microsoft Office, like Calibri, Arial or Times New Roman. Other fonts become the closest common one: Arial for plain fonts, Times New Roman for fonts with serifs (the small strokes at the ends of letters) and Courier New for typewriter fonts. Bold and italic are kept. Because the letters can be a little wider or narrower, a page of text can run onto an extra page: our 117-page maths textbook gave 122 pages in LibreOffice.
• Maths keeps its letters, not its drawing. Formulas come across as letters, numbers and symbols, but fraction bars, root signs and tall brackets are drawn lines in a PDF and are left out, and parts of a formula can land on lines of their own. Word's equation editor isn't used, so formulas need tidying by hand.
• Shapes and backgrounds are left out. Coloured boxes behind text, underlines, highlighting and lines between columns are drawn shapes in a PDF, not part of the text. Drawings made of many lines or curves, like charts and diagrams, come across as pictures.
• Words inside pictures stay in the picture. The labels of a chart or a diagram are part of its picture, so they can't be edited in Word.
• Repeated headers and footers are left out. A line that repeats at the top or bottom of every page, like a book's title or a chapter name, is taken out of the text. Page numbers become Word's own page numbers instead.
• Footnotes stay as small text at the foot of their page or column, not as Word footnotes.
• Sideways text. A label set at an angle, like a word running up the edge of a page, goes at the end of its page as ordinary text. A page whose text all runs sideways or top to bottom, like vertical Japanese, goes in as a picture, and the result says so.
• Text the PDF can't give back. Some PDFs use fonts that don't say which letters they draw, so their text would come out as nonsense. Those pages go in as pictures, and the result says which ones. In some PDFs in Hindi and other Indian scripts, letters that join others, like vowel signs, aren't stored one by one, and pdf.js can't read them back in full. In two Hindi PDFs we made to test this, one saved from LibreOffice and one from Chrome, many words lost vowel signs or joined letters. When this happens the result warns you, so check the text against the PDF.

Scanned PDFs

A scan is a photo of paper. Its pages hold pictures of words, not words, so there is no text in them to copy into Word. Reading the words in a picture is called OCR, short for optical character recognition, and this converter doesn't do OCR.

What you get instead: each scanned page goes into the Word file as one picture, on a page of the same size, sharp enough to read and print (150 ppi). It looks right, but you can't edit its words. The result says plainly that the PDF is a scan, or which of its pages are, for example "Pages 3 and 7 are scanned". A PDF with typed pages and a few scanned ones gives text for the typed pages and pictures for the scanned ones.

Scans that already have their text: many scanner apps and office copiers run OCR themselves and hide the recognised text behind the picture, so the PDF can be searched. When a PDF has such a hidden text layer, the Word file gets that text instead of the picture, and the result tells you, because OCR makes mistakes: check names, numbers and amounts.

Is my PDF a scan? Open it and try to select a single word. If you can't, or the whole page is selected like a photo, it is a scan.

If you need editable text from a scan, you need an app that does OCR. Many of them, including Google Docs when you open a PDF from Google Drive, read the text on their own servers, which means uploading your document.

Real results

We converted these files in Chrome on our test computer. The time is the one the page showed when it finished, from pressing Convert to the finished Word file. Your device may be faster or slower, and phones usually take longer. We then opened every Word file in LibreOffice to count its pages, and looked for each word of the PDF in it.

• 99.5%of our research paper's words found in its Word file
• 14 pagesin LibreOffice for the 14-page, two-column paper
• 2.8 sfor the 14-page paper
• 0 bytessent to any server

Pages of text are quick, and the Word file is usually smaller than the PDF, because it doesn't carry the PDF's fonts. Scans take longer and make bigger files, because every page is a picture.

How we counted words: we read each PDF's words with pdftotext, a different PDF reader, and looked for each one in the Word file. Most of the few we didn't find were labels inside charts, which come across inside the chart's picture, bits of formulas, and words that pdftotext itself reads in two pieces.

Editing the Word file

The Word file is a .docx, the format Word has used since 2007, so Microsoft Word, Google Docs, WPS Office, LibreOffice and Apple Pages can all open it, on a computer or a phone. A few things make editing easier:

• Turn a table's rows into a Word table. In Word, select the rows, then choose Insert, Table, Convert Text to Table, and pick Tabs as the separator. LibreOffice has the same under Table, Convert, Text to Table.
• Jump between headings. In Word, turn on View, Navigation Pane to see every heading and jump to it. In Google Docs, the outline at the side does the same.
• Let text run from page to page. Each PDF page starts a new page in Word, so text you add on one page pushes only that page's text down. To join two pages, show the formatting marks (the ¶ button on Word's Home tab) and delete the section break at the end of the first page.
• Change a font everywhere. Select all the text (Ctrl+A, or Cmd+A on a Mac) and pick a font and size. Headings keep their Heading 1, 2 and 3 styles, so they stay in the navigation pane.
• Page numbers look after themselves. They are Word's own, so they stay right when pages are added or taken away. Double-click the footer to change them.

PDF to Word on Android, iPhone, Windows and Mac

The converter runs in your web browser, so there is nothing to install: no app, no extension and no account.

• Android

  Open this page in Chrome, Samsung Internet or another up-to-date browser and tap Choose PDF files. The Word file goes to your Downloads folder. Open it with the Word, Google Docs or WPS Office app.

• iPhone and iPad

  Open this page in Safari or Chrome and pick a PDF from the Files app. Downloads usually go to Downloads in the Files app. Tap the Word file there and share it to Word, Google Docs or Pages to edit it.

• Windows, Mac, Linux, Chromebook

  Use Chrome, Edge, Firefox or Safari. Drop PDFs onto the page, or choose them. The Word files go to your Downloads folder.

Long PDFs on a phone. Pages are read one at a time, and only the finished Word file is kept, so long PDFs work. Scans need the most memory: in our test on a computer, the browser used 1,301 MB at most while converting the 400-page scan, up from 527 MB before. Keep the page open with the screen on until it finishes, because browsers slow down or pause pages you aren't looking at.

Private by design

Most online PDF to Word converters work by uploading. Your PDF goes to their server, it is converted there, and you download the Word file back. For that time your document sits on a computer you don't control, and you have to trust that it is deleted afterwards.

NoUploadPDF works the other way round. Your own browser reads the PDF and writes the Word file, on your phone or computer. Your PDF and the Word file never leave your device, and the page is not allowed to send data to any other website. That matters for a CV, a bank statement, a contract, a medical report, or papers like an Aadhaar or PAN card.

Passwords stay on your device too. When a PDF is locked, you type its password on the PDF's own row, and it is used only to open the file here.

You can check all this yourself: open your browser's network panel while converting, and you'll see that nothing is uploaded. Our privacy policy has the details.

When something goes wrong

• "Can't be opened: it may be damaged, or not a PDF." The file may be broken, cut short while downloading, or another kind of file with a .pdf name. Try opening it in a PDF reader. If that works, save or print it as a new PDF there, and add the new copy.
• It asks for a password. The PDF is locked. Type the password on its row and press Open. Until then it is left out, and your other PDFs are converted without it.
• The pages came out as pictures. The PDF is a scan, or its text can't be read back. The note under the Word file says which. See Scanned PDFs.
• The Word file has more pages than the PDF. Word's fonts are a little different from the PDF's, so a full page can run over. Make the font size a little smaller, or the margins a little narrower, in Word.
• A few symbols show as empty boxes. The app you opened the file in has no font with those symbols. Try another app, or a computer.
• Hindi or another Indian script looks wrong. Many of these PDFs don't store every letter in a way that can be read back, and the result warns you when that happens. Compare with the PDF, and fix the words by hand.
• It's slow. Long scans take the most time: our 42-page scan took 14.6 s on a computer. Keep the page open in front until it finishes.
• The download didn't show up. Look in your Downloads folder, or in Downloads in the Files app on an iPhone. If it isn't there, press the Download button again: your Word files stay ready until you change something or leave the page.

Tips for the best Word files

1. Start from the best PDF you have. A PDF saved from Word or Google Docs converts better than a printout that was scanned, because it has real text.
2. For a CV or a letter, convert it, then check the fonts and line spacing in Word before you send it. Select all the text to change its font at once.
3. For tables, turn the rows into a Word table with Convert Text to Table, as described in Editing the Word file.
4. Check the numbers in anything that came from a scan's hidden text, because OCR mistakes are easy to miss: a 5 read as an S, or a missing decimal point.
5. Only need a few pages? Take them out first with Split PDF, then convert the smaller PDF.
6. Need pictures, not text? PDF to JPG saves each page as a sharp picture, which is better for posting or sharing a page as it looks.
7. Keep your PDF. It is the original, with the exact look, fonts and layout. Keep it next to the Word file.

DOCX, OCR and other terms, explained

DOCX

The file format of Word documents since Word 2007. Most document apps can open and save it.

Scan

A PDF made from photos of paper. Its pages are pictures, with no text in them.

OCR

Optical character recognition: software that reads the letters in a picture and turns them into text.

Text layer

Hidden text behind a scanned page, put there by OCR so the PDF can be searched.

Heading style

Word's way of marking a title, like Heading 1 for a chapter and Heading 2 for a section. It builds the navigation pane and tables of contents.

Tab stop

A set place along a line that text jumps to when you press Tab. It lines up columns without a table.

Section break

A mark in a Word file where a new part starts, with its own page size, margins or columns.

ZIP

One file that holds many files. Your phone or computer can open it and take the files out.

Limits

• No OCR: scanned pages become pictures. A scan with a hidden text layer gives that text.
• Tables become rows with tab stops, not Word tables, and their lines and shading are left out.
• Formulas keep their letters and symbols but lose fraction bars, root signs and tall brackets.
• The PDF's fonts are swapped for common Office fonts, so text can take a little more or less room than in the PDF.
• Each PDF page starts a new Word page; text doesn't flow from one page to the next until you join them.
• Words inside charts and diagrams stay part of their picture.
• Pages whose text runs sideways, or whose text can't be read back, become pictures. In some Hindi and other Indian-script PDFs, letters can be missing.
• Long scans take time and memory: our 400-page scan took 145.6 s on a computer. Keep the page open until it finishes.
• A PDF that can't be opened, or a page that can't be read, is left out with a note. Everything else is still converted.