PDF to Word Converter

A PDF is built for printing, not editing: every character sits at a fixed spot on the page, and there is no such thing as a paragraph inside the file, only lines of text placed at coordinates. This converter reads those coordinates, along with each piece of text's font, size, color, angle, and any link attached to it, and writes a real Word document (.docx) from them. Every PDF page becomes its own page in Word with the same paper size and orientation, so page 3 of the PDF is still page 3 of the Word file, and a landscape page stays landscape.

There are two layouts, because "looks identical" and "easy to rewrite" pull in different directions. Exact layout is the default and keeps the look of the page: the lines, table borders, colored boxes, logos, and photos are placed on the page as one picture behind the text, and every line of text sits on top of it in its own invisible text box at its original position, in the same font, size, and color. Rotated labels stay rotated, and superscripts and subscripts keep their raised or lowered position. Editable text instead rebuilds real paragraphs that rewrap as you type: it works out which lines belong together, keeps left, centered, right, and justified alignment, first-line and hanging indents, the spacing between paragraphs, and turns rows of columns (a price list, a simple table) into tab-aligned lines. A page set in two or three columns, like a newsletter or a journal article, gets real Word columns, with a title that ran across the top kept full width above them, and text that sat on a colored box or a dark page keeps that color as paragraph shading so it stays readable. Photos are cropped out and placed between the paragraphs where they appeared, links stay clickable, and larger headings are marked so they show up in Word's navigation pane.

What makes Exact layout possible is how each page is read: the converter draws the page once with every text glyph caught just before it is painted, which yields both a picture of everything except the text and the exact spot and color of each glyph the PDF actually shows. That also means text the PDF hides, such as an invisible OCR layer under a scanned page, is never pasted on top of the scan as a second copy. A page that is only a scan has no text to extract, so it is kept as a picture; this tool does not perform OCR, and says so after converting. Text in fonts the PDF never mapped to real characters, which would come out as gibberish, is kept as part of the page picture instead.

All of this runs in your browser using Mozilla's PDF.js to read the file and a small built-in writer to assemble the .docx, so your document is never uploaded, and the page works offline once it has loaded. The result opens in Microsoft Word, LibreOffice Writer, Pages, and other editors that read .docx. Word uses the fonts installed on the computer that opens the file, so if the PDF used a font you don't have, Word substitutes a similar one and line lengths may shift slightly; in Exact layout the converter measures the font in your browser and adjusts the spacing so each line keeps its original width, widening only the word gaps on justified lines so their right edges stay straight.