Convert Word to PDF in your browser
Your DOCX is read and drawn as a PDF here, with no upload and no account. It is a rendering rather than the file Word would export, so read the note about fonts before you send it.
A four page report with three photographs in it, so you can see for yourself what this keeps and where it stops matching Word.
or browse your device. Everything is processed on this device
No DOCX to hand?
The PDF is mostly the pictures inside your .docx, embedded as they are. Shrinking those in the document first makes both files smaller, and the text is untouched.
Compress the Word file firstHow to convert Word to PDF
It is unzipped and read in this tab. There is no upload endpoint in this site for a contract, an application or a report to go to.
Keep the document size unless it has to print on different paper. Choose Times Roman if your document is set in a serif face, Helvetica if it is not.
The result panel says what was drawn, what was substituted and what could not be carried. It is worth ten seconds before the file leaves your hands.
Line breaks and page breaks will not match Word exactly. If the document is going somewhere that cares about that, export it from Word instead.
What this reads, what it invents, and where it stops
A .docx is a ZIP archive with the document model inside it. word/document.xml holds the paragraphs, the runs of formatting inside them, which style each one uses and where every picture goes; styles.xml holds what those styles mean; numbering.xml holds the lists. All of that is read here rather than guessed at, which is why this conversion is possible at all and why the reverse one is not. A PDF has no paragraphs to find, which is what the PDF to text page here explains at length.
What a .docx does not contain is the layout. Word decides where a line breaks and where a page ends at the moment it draws, using the font metrics of the machine it is running on, which is why the same file paginates differently on two computers. It also does not contain its fonts: Calibri, Aptos and Cambria are named by the document and installed on the computer, not stored in the file. A PDF reader is only guaranteed to carry fourteen fonts, and none of them are those. So the text here is measured and drawn in Helvetica or Times, every line is broken again from scratch, and the result is a rendering of your document rather than a copy of it.
In practice that means the words, the order, the headings, the bold and italic, the colours, the alignment, the indents, the list numbering, the table columns and the photographs all come through. What moves is where the lines break, which paragraph a page ends on, and therefore how many pages there are. A four page report can come out as three or five, and nothing has gone wrong when it does.
Pictures are the good news. At full quality a photograph stored as a JPEG goes into the PDF byte for byte, because PDF has a native JPEG stream, so it is exactly as sharp as it was and nothing is re-encoded. A PNG is packed again losslessly, every pixel intact. A GIF, a BMP, a TIFF, a WebP or an SVG cannot be carried as it stands, because a PDF stream holds a JPEG or a PNG and nothing else, so each one is decoded here and written into the page again as a lossless PNG. At the picture quality this page arrives at, nothing is re-compressed with loss on the way, so what lands in the PDF is what came out of the decoder, which does mean a big scan makes a big PDF. Lower that setting and every opaque picture is written as JPEG at the quality you chose, the photographs the document already held included. That is the whole point of the control, and on a document that is mostly photographs it is the only thing that moves the size at all. Three of those formats are decoded by this site's own code; WebP and SVG are decoded by your browser, which every current one can do.
Three of those conversions involve a choice, and the result note names whichever of them your document ran into. An animated GIF is drawn as its first frame, because a page in a PDF does not move. A TIFF holding several pages is drawn as page one, because a picture in a document is one image. An SVG is a drawing rather than pixels, so it is drawn at twice the size it declares and embedded as that picture: sharp at any sensible zoom, but no longer resolution independent the way the original was.
What still has nowhere to go is a vector metafile. A pasted Excel chart, a SmartArt diagram or an equation is stored inside the .docx as an EMF or WMF drawing, which is a list of instructions for the Windows drawing system rather than a picture, and neither this tool nor your browser has anything that can follow them. Those two formats, and only those two, are drawn as an empty outline the right size and counted in the note, so the hole is visible rather than silent.
A list of what is not carried, because a page that leaves this to be discovered is worse than one that says it: headers, footers, footnotes, endnotes, comments, text boxes, shapes, charts, cell shading, vertically merged table cells, tab stops, and fields such as page numbers or a table of contents. Documents written in Chinese, Japanese, Korean, Cyrillic, Greek, Hebrew, Arabic or any other script those fourteen fonts do not cover are refused outright, with the reason, rather than converted into pages of question marks.
If you have Word, Pages or LibreOffice in front of you, their own export is better than this and always will be, because they have your fonts and this does not. This page is for the times when you do not have them: a borrowed machine, a phone, a locked down work laptop, or a document you are not willing to upload to a stranger to convert.
Word to PDF FAQ
Will the PDF look exactly like my Word document?
No, and no browser based converter can promise that. Your document names fonts such as Calibri or Cambria without containing them, and a PDF reader only guarantees fourteen fonts, none of which are those. So the text is drawn in Helvetica or Times, which measure differently, which means lines break at different words and pages end at different paragraphs. Everything else, the words, the order, the headings, the bold, the colours, the lists, the tables and the photographs, comes through. If a document has to match to the millimetre, export it from Word itself.
Why did my four page document come out as three pages?
Because the pages were worked out again from scratch. A .docx does not store where its pages end, only its text and styles, and Word decides the rest as it draws using fonts this tool does not have. With a slightly narrower or wider substitute, the same words need a different number of lines and therefore a different number of pages. The result panel tells you how many pages came out so you are never surprised by it later.
Are my photographs kept, and do they still carry their EXIF and GPS?
Yes to both at the quality this page arrives at, and the second half matters. A JPEG inside the document is embedded in the PDF byte for byte with no re-encoding, so it is exactly as sharp as it was, and everything the camera wrote into it travels too, which on a phone photo usually includes GPS coordinates, the date and the device. If the PDF is going to a landlord, an employer or a public form, that is worth knowing. Lowering the picture quality changes that answer: a re-encoded picture is rebuilt from pixels, so the EXIF and the GPS go with the bytes they were attached to. If stripping them is what you want, that is the way to do it here. The properties Word keeps about the document itself, the author, the company and the revision history, are a separate thing and are not carried across at all, because the PDF is written from nothing.
What happens to tables, bullet points and numbered lists?
They are laid out. Table columns use the widths in your document, cells wrap their own text, rows are never split across a page, and cells merged sideways are honoured. Numbered lists keep their counters, including multi level ones such as 2.3.1, and a bulleted list gets an ordinary round bullet, because Word draws its own in Symbol and Wingdings and no standard PDF font carries either. What is not carried is cell shading, table styles, borders that differ per side, and cells merged vertically, which are drawn empty.
My chart or diagram is missing. Where did it go?
A pasted Excel chart, a SmartArt diagram or an equation is not a picture inside a .docx: it is an EMF or WMF vector drawing, which is a list of drawing instructions for Windows rather than an image, and nothing here or in your browser can follow them. Those two formats are the only ones still left out. Rather than reflowing the page as though the drawing was never there, an empty outline of the right size is drawn in its place and the result note counts them. To include one, screenshot it, paste the screenshot into the document in place of the drawing, and convert again.
What about GIF, BMP, TIFF, WebP or SVG images in my document?
All five are drawn. A PDF stream can hold a JPEG or a PNG and nothing else, so each of those is decoded and written into the page again as a lossless PNG, unless you lower the picture quality, at which point the opaque ones go in as JPEG at the quality you chose, exactly as the photographs do. The three choices that involves are stated in the note after the conversion: an animated GIF is drawn as its first frame, a TIFF with several pages is drawn as page one, and an SVG is a drawing rather than pixels so it is rasterised at twice its declared size and is no longer resolution independent. WebP and SVG are decoded by your browser rather than by us, so a very old browser may still show an outline for those two.
Can it convert a document written in Chinese, Japanese or Russian?
No, and it says so rather than trying. The fonts built into every PDF reader cover Western European characters and nothing else, and embedding a font that covers more would mean shipping several megabytes of typeface to every visitor. A document mostly in one of those scripts is refused with that explanation. A document in English with a few symbols in it converts normally, and the note counts any character that had to be replaced.
What about headers, footers, page numbers and footnotes?
None of them are drawn. They live in separate parts of the file with their own layout rules, and a header drawn half right is worse than an absent one. Page numbers and a table of contents are fields, which means Word calculates them rather than storing them, and the numbers it calculated are for its own pagination rather than this one. The note names whichever of these your document actually contains, so you know what is missing rather than having to spot it.
Can it open an old .doc file?
No. The old binary .doc is a completely different format that happens to share three letters of its name: it is not a ZIP and none of this applies to it. Open it in Word or LibreOffice and save it as .docx first. The tool recognises one and tells you that rather than reporting a damaged file.
Are my DOCX files uploaded to convert them?
No. The DOCX is decoded and the PDF is written by your own browser, in this tab. There is no upload endpoint in the site for a file to go to, which you can check in your network tools: convert something and watch that no request carries it. Disconnect from the network after the page loads and it still works.
Related
Same format joins the queue. Anything else re-routes to its own tool.