Make a Wikipedia book: several articles in one PDF

Collect the Wikipedia articles you want, put them in the order you want them, and get a single PDF with a cover, a contents page and a chapter for each article. It runs in your browser: no account, no sign-up, no install.

Open the app

Wikipedia’s own book tool is gone

If you have gone looking for this, you have probably found the wreckage of the Book Creator. Wikipedia used to let you gather articles into a “book” and download the result. The book rendering service was deprecated in 2021, the Book namespace was uninstalled from the English Wikipedia, and the saved books were deleted along with it.

A single article still has a Download as PDF link in Wikipedia’s sidebar. It is the several-articles-in-one-file part that went away, and it has not come back.

That is the part PediaPDF does. It is not a restoration of the old tool and it is not run by Wikipedia — it is a separate, free thing that reads the same articles and typesets them into one document on your own machine.

How to make one

  1. Open the app and switch the toggle above the search box from Article to Book. A basket appears beside the results.
  2. Search for an article as normal, then press + Add on the result you want. The button changes to ✓ Added and the article drops into the basket. Search again and add the next one — the basket keeps everything you have collected, and the count sits in its header.
  3. Put them in order. Drag a row, or use the and buttons beside it. The order in the basket is the order of the chapters.
  4. Press Create book. The articles are fetched together, typeset as one document, and the finished PDF opens in the viewer, where you can still change the fonts, paper size and margins and watch it re-set.

Already have a list? Press Edit as text in the basket header and paste it in — one article name or one full Wikipedia URL per line. Press Done and they all land in the basket. This is also the quickest way to see what you have collected as plain text, or to move a list between devices.

Your basket survives leaving the page, so you can build a book over several sittings. Books you have made appear in the app’s Recent list, and picking one there puts its articles back in the basket rather than regenerating blindly — so a book is something you can come back to and amend.

What the PDF looks like

It is a book, not a stack of separate exports glued together. One document, one set of page numbers, one contents page.

Leaving things out

A book of ten articles is long, and most of it is probably not what you wanted. Before generating, you can open each article in the panel and untick what you do not need — a whole chapter, a single section, the references, the “See also” list. What is left is what gets typeset, and the contents page and bookmarks are built from what survives.

If you drop an article entirely, it also leaves the cover and the document title. The book describes itself as it actually is. The same panel does finer work than whole sections — see choosing which sections to keep.

Articles from more than one language

A book does not have to be in one language. Search in whichever Wikipedia you like and add from it, or press Edit as text and paste full URLs from different editions — https://de.wikipedia.org/wiki/Bach and https://fr.wikipedia.org/wiki/Bach in the same list are fine. Each article is fetched from its own Wikipedia, and each chapter is typeset with the fonts and hyphenation its own language needs.

One honest limit: the document’s own furniture — the word “Contents”, the footer lines — is written in a single language, taken from the first article. The chapters are right; the frame around them is not per-chapter.

What to expect of it

Everything happens on your machine, which is what makes it free and private, and also what sets the ceiling. There is no limit on how many articles a book may hold, but a twenty-article book is twenty articles’ worth of downloading, parsing, image fetching and typesetting, and a phone will feel that where a laptop will not. Start with a handful.

Arabic and Hebrew are worth a word. They are shaped and printed correctly, but as images rather than as text, so that part of a book is not selectable or searchable. Latin, Greek, Cyrillic, Indic and CJK scripts come out as real text. There is a fuller account of what does and does not work well on the about page, along with exactly what leaves your browser.

If the book is going on paper, the paper size, the margins and the picture quality are all worth a moment first — see printing a Wikipedia article.

The text is Wikipedia’s and stays under CC BY-SA 4.0; the cover carries that attribution for you. You may keep the PDF, print it, or pass it on.

Make a book

Free, and nothing to sign up for.

Last updated 3 September 2026.