Convert DOCX to PDF

1. 1

   Add your document

   Drop a DOCX, DOCM, DOTX, DOTM, ODT, RTF or TXT file onto the upload area, or click Browse. The file is read on your device and never uploaded. If a format cannot be converted, the tool says which format it is and what to do instead rather than failing part way through.

2. 2

   Choose the page setup

   Leave everything on Auto to keep the page size, orientation and margins the document already has. Change them when you need a different sheet — an A4 report sent to a US office becomes Letter with one press, and the text re-flows to fit rather than being scaled down.

3. 3

   Convert

   Press Convert to PDF. The document is parsed, laid out against real font metrics and written as text runs, so the result is searchable and copyable. Progress is shown per file, and long documents keep the page responsive because the work runs off the main thread.

4. 4

   Check the preview and download

   The finished PDF opens in the preview below the queue. Check the page breaks, the tables and the links, then download. If you change a setting afterwards, finished results are marked out of date and one press re-applies the new setting to all of them.

Most free Word to PDF converters upload your document to a server, meter you at two files a day, and hand back a PDF you cannot search. This one runs on your device and writes real text.

• Selectable, searchable text: the PDF contains the document's characters, not a picture of them, so Ctrl+F works, copy works, and a screen reader can read it.
• Your own page setup: page width, height, orientation and margins are read from the document, per section, so a Letter file stays Letter and a landscape appendix stays landscape.
• Links that still work: every hyperlink becomes a real clickable annotation in the PDF, and adjacent words of one link share one clickable area.
• A navigable outline: headings become PDF bookmarks, nested to nine levels, so a long report opens with a contents panel instead of nothing.
• Seven input formats: DOCX, DOCM, DOTX and DOTM, plus ODT from LibreOffice, RTF and plain TXT.
• An honest report: anything the converter could not carry across — a vector drawing, a chart with no preview, a substituted font — is listed under the file rather than quietly left out.
• Batch and merge on Pro: up to 50 documents at once, delivered as one ZIP or as a single merged PDF in queue order.

The single most important choice on this page is the first one in the settings, and it decides what kind of PDF you get. Selectable text is the default and is what you want almost every time; flattening is a deliberate trade you make for a specific reason.

Selectable text — the default

• It is searchable. Ctrl+F inside the PDF finds words. So does your operating system's file search, and so does any document management system the file is filed into.
• It is copyable. A quote can be lifted out of the PDF and pasted somewhere else with its spelling intact, rather than retyped from a picture.
• It is readable by machines. Applicant tracking systems parse CVs as text. A CV converted to page images parses as nothing at all, which is a genuinely expensive way to lose an application.
• It is accessible. A screen reader reads text. It cannot read a picture of text, and no amount of resolution changes that.
• It is small and stays sharp. A five-page document is around 12 KB rather than a megabyte and a half, and it prints crisply at any size because the letters are outlines, not pixels.

Flattened image — when it is the right answer

• The layout must not move. A signed form, a price list, a certificate — anything where you would rather guarantee the appearance than allow the text to be lifted.
• The document is in a right-to-left or Indic script. Arabic, Hebrew, Devanagari, Thai and their relatives need contextual shaping and bidirectional reordering. Flattening renders the page properly instead of guessing.
• You want the text not to be extractable. Flattening is not encryption and it is not security, but it does stop casual copy-and-paste.

The cost of flattening is stated rather than hidden: the file is many times larger, nothing in it can be searched or copied, and it blurs when zoomed past its rendering resolution. If you are unsure, convert once each way and compare — both take seconds.

A Word document is an editable working file whose appearance depends on the machine that opens it. A PDF is a finished one whose appearance does not. That difference is the whole reason this conversion exists, and it is why almost every organisation asks for PDF at the point where a document stops being drafted and starts being sent.

Benefits of PDF Format

• It looks the same everywhere: a PDF carries its own page geometry, so it opens identically on a phone, a Mac, a Windows laptop and a printer's RIP. A DOCX re-flows to whatever fonts and page setup the reader happens to have.
• Nothing gets edited by accident: a reviewer cannot leave a stray keystroke in a PDF the way they can in a document that opens in edit mode.
• It is what forms and portals accept: job applications, tender submissions, university coursework and government filings almost universally specify PDF and often reject anything else.
• It is far smaller: a text PDF drops the revision history, the style definitions, the embedded thumbnails and the editing metadata a DOCX carries around.
• It archives cleanly: PDF is a published, stable specification. A document saved as PDF today opens in twenty years; a proprietary editor format is a bet on that editor still existing.

A Word document does not have one page setup — it has one per section, and a well-built report uses that. The cover page may be landscape, the appendix a different paper size, the body A4 with mirrored margins for binding. This converter reads all of it, section by section, and gives each section's pages their own geometry in the PDF.

When you should override it

• Sending an A4 document to a US office: pick Letter. The text re-flows onto the shorter, wider sheet instead of being scaled down, so the type stays the size the author chose.
• A wide table that will not fit: force Landscape. Columns that were being squeezed get their width back.
• Saving paper on a draft: Narrow margins fit noticeably more text per page; None removes them entirely for a document that is going to be trimmed or bound edge to edge.
• A poster or a plan: A3 gives you a sheet twice the area of A4 at the same proportions, so nothing needs re-designing.

Overriding the page size re-flows the text; it never scales the page down to fit. That distinction matters because scaling is what makes a converted document arrive with type too small to read comfortably, and it is what a converter does when it cannot re-lay-out the content.

A PDF either carries the shapes of the letters it uses or names a font it expects the reader to have. This converter does both, choosing per character, which is why an English document comes out at a few kilobytes and a Polish or Greek one is still small.

How a font is chosen

• Arial, Helvetica, Times New Roman and Courier New: matched exactly by metrics using the fonts every PDF reader already has, so nothing is embedded and the line breaks land where the author saw them.
• Other Latin fonts: mapped to the closest of those three by shape — serif, sans or monospaced. Calibri, the modern Word default, is about 7% narrower than Arial, so line breaks can differ slightly; the tool names every substituted font under the file.
• Accented, Central European, Cyrillic and Greek text: drawn from an embedded face, subset to only the characters the document actually uses, so a Cyrillic paragraph in an English report costs a few kilobytes rather than a font file.
• Chinese, Japanese and Korean: drawn from an embedded CJK face, also subset. A page of Chinese adds roughly 50 KB rather than the 13 MB an unsubset face would.

Where a character has no glyph in any available face, it is replaced visibly and counted in the file's notes, rather than dropped. A missing character that disappears silently is worse than one you can see, because you cannot proofread what is not there.

A DOCX file is a ZIP archive of XML parts. Converting it properly means reading those parts, resolving what the document actually says, laying it out, and writing the result — four steps, each of which a shortcut converter skips.

Conversion Process

• Unpack and resolve: the archive is opened and its parts are located through the package's own relationship map rather than by guessing at file names, so documents written by Google Docs, Pages and template engines are read as reliably as ones written by Word.
• Resolve the styles: every paragraph's appearance comes from a chain — document defaults, then the default style, then the paragraph's style and everything it is based on, then the list level, then its own direct formatting. Applying those in the wrong order is what makes headings lose their numbering indent.
• Build the document: paragraphs, runs, tables, lists, images, headers and footers become one structure with every measurement already converted to points, so nothing downstream has to know that Word counts in twentieths of a point and pictures in 914,400ths of an inch.
• Lay it out: lines are broken against real font metrics, paragraphs are placed with widow and orphan control, headings are kept with the text that follows them, and table rows break between rows rather than through them.
• Write the PDF: each line becomes text-showing instructions, hyperlinks become link annotations, and headings become an outline. Adjacent pieces of the same line are merged into one instruction, which is what keeps the file small and the text extractable as whole words.

Text costs almost nothing in a PDF. Pictures are the entire story of how big the file ends up, and the Image quality setting is the one control that changes it materially.

What each quality setting does

• Original: pictures are copied into the PDF untouched, byte for byte. Nothing is decoded and nothing is re-compressed, so a photograph loses nothing at all — and the PDF is as large as the pictures in it.
• Print (300 dpi): sized for commercial printing. A photo placed three inches wide keeps 900 pixels across, which is more than any press needs and far more than any screen.
• High (200 dpi) and Balanced (150 dpi): sized for an office printer and for reading. 150 dpi is indistinguishable from the original at normal reading distance and is usually where the largest saving comes from.
• Small (96 dpi): screen resolution, for a document that will only ever be read on a display or sent by email.

Downsampling is measured against the size a picture is actually printed at, not its pixel count. A 4000-pixel photo placed three inches wide is carrying 1,333 dpi; at Balanced it needs 450 pixels and the other 3,550 columns are bytes nobody will ever see. Pictures with transparency are kept as PNG so a logo does not gain a black box; everything else becomes JPEG, and a re-compression that would have made a file bigger is skipped.

Some numbers from this converter, measured rather than estimated, on a desktop machine with an ordinary document.

Key Statistics

• A five-page document with tables and hyperlinks converts in about 16 milliseconds and produces a 12 KB PDF.
• A 306-page document converts in about 1.1 seconds — roughly 3.7 milliseconds per page, and the cost per page does not grow with length.
• Twenty documents converted as a batch finish in well under a second, spread across several background threads so the page stays responsive.
• An English document embeds zero font bytes, because the fonts it needs are the ones every PDF reader already has.
• A page of Chinese adds about 50 KB of subset font — the same page with the font embedded whole would add over 13 MB.
• Every character in the source appears in the PDF: measured at 100% against a text extractor that reads the document independently of the converter.

A few habits make the difference between a PDF that is merely produced and one that is genuinely finished.

Quality Considerations

• Use real heading styles in Word rather than manually enlarged bold text. Headings are what become the PDF outline, and a document styled by hand produces no outline at all.
• Check the notes under each converted file. They name every substituted font and every object that could not be carried across, which is the fastest proofread available.
• Leave page setup on Auto unless you have a reason to change it. The document's own geometry is almost always what the author intended.
• Set Document properties to Strip before publishing anything. A Word file carries the author's account name and often every reviser's.
• Choose Original image quality for anything going to print, and Balanced for anything going by email. The difference in file size is usually an order of magnitude.
• Open the preview and scroll it before downloading. Page breaks are the one thing worth a second of your attention, especially in a document with long tables.

Common Use Cases

• CVs and job applications: applicant tracking systems read text. A CV flattened to page images parses as an empty document, which is the most costly possible version of this mistake.
• Contracts and agreements: a PDF fixes the wording and the pagination, so "clause 4 on page 3" means the same thing to everyone.
• Coursework and dissertations: a submission portal that specifies PDF usually also specifies a page size and margins, and both are settings here rather than something to fight with in Word.
• Reports and proposals: a heading outline turns a sixty-page document into something a reader can navigate instead of scroll.
• Invoices and statements: a searchable PDF can be found later by amount or reference number; a scanned-looking one cannot.
• Documents from LibreOffice: ODT converts directly, so there is no round trip through Word and no formatting lost on the way.

The interesting part of this conversion is the layout, because a Word document does not describe pages — it describes a stream of content and the rules for arranging it. Turning that into fixed pages is the actual work.

Conversion Process

• Line breaking: each paragraph is broken into lines using the exact advance width of every character at its exact size. Justified text distributes the leftover space between words; a word too long for the column is broken inside rather than allowed to run off the page.
• Line height: single spacing means the font's own ascent, descent and line gap — about 1.15 times the point size for Arial — not a round number. Exact, at-least and multiple spacing are three different rules and are treated as three different rules.
• Page breaking: lines are placed until the page is full, then a break is chosen. A heading marked keep-with-next moves forward with its paragraph, and widow and orphan control stops a single line being stranded on its own page.
• Tables: column widths are resolved against the available text column and scaled proportionally if the table is wider, rather than being clipped. Rows break between rows, header rows repeat at the top of every page the table continues onto, and vertically merged cells keep their merge.
• Headers, footers and page numbers: these are laid out once per page, because a footer containing a page number is different on every page. A converter that lays them out once prints page 1 on all of them.
• Writing it out: every positioned line becomes text instructions in the page's content stream, with adjacent pieces merged so that a reader reconstructs whole words rather than isolated glyphs.

A five-page document converts in about 16 milliseconds on a desktop machine and a 306-page one in about 1.1 seconds, because nothing is rasterised: the pages are written as text.

Seven input formats, and one that is refused on purpose. The list is enforced when you pick the file, by reading its first bytes rather than trusting its name — so a file with the wrong extension is identified correctly instead of failing half way through a conversion.

• DOCX: the standard Word format since 2007, and the one most documents are in.
• DOCM: a macro-enabled Word document. The content converts normally; macros are not code this tool runs, so they are simply absent from the PDF.
• DOTX and DOTM: Word templates. They have the same internal structure as a document and convert the same way, which is useful for checking how a template actually prints.
• ODT: the OpenDocument format written by LibreOffice, OpenOffice and Google Docs exports. Styles, lists, tables and images all carry across.
• RTF: Rich Text Format, still produced by older applications and by many database and reporting systems. Formatting, colours, tables and page setup are read.
• TXT: plain text, laid out in a monospaced face with tabs kept as real tab stops so aligned columns stay aligned. UTF-8, UTF-16 and Windows-1252 files are all detected correctly.
• DOC — refused, and here is why: the pre-2007 binary format stores text in fragments and a separate table that says how to reassemble them. A partial reader produces paragraphs in the right order with the wrong words in them, which looks convincing and is wrong. Open it in Word or Google Docs, save as DOCX, and convert that.

• Word compressor
• Word merger
• Word viewer
• Word to text
• Word to HTML
• PDF merger