Text to PDF

It is not: this tab assembles the PDF and the text stays on your device. Apart from the page's own code, the only thing it may request from pdf.xyz is font files, the Geist Mono monospace face, a Noto face, or both, when your text or file name needs characters beyond Western European. Each request asks for a face by its file name, so it reveals which kind of face was needed, but none of your words.

Up to 2 MB, whether you pick a file or paste. A file larger than that on disk is turned away when you load it, and the text is measured again in UTF-8 bytes when you press Create PDF; anything over is refused with a message rather than half-converted. A Chinese or Bengali character takes several bytes in UTF-8, so such text reaches the cap at far fewer characters than English. Empty or whitespace-only text is refused too.

Beyond Western European characters, text in Bengali, Latin, Greek, Cyrillic, Devanagari, Hiragana, Katakana, Kanji, Hangul and Han gets a bundled face embedded in the file: the Geist Mono monospace face where it can draw every character, as for Polish or Russian, and otherwise a Noto face. Arabic and Hebrew stop the conversion with an explanation: this tool places each line from the left margin rightward, so a right-to-left line would read in reverse. Emoji stop it too, because colour emoji are not embedded, so delete them before converting. Any other script without a bundled font is refused with the offending characters shown; the converter never swaps a character it cannot draw for a placeholder. One document cannot mix scripts that need different fonts, and the file name counts as part of the document here, so put, say, Chinese and Bengali passages in separate files.

There is none: every PDF is US Letter with fixed margins and text size, and the page heading doubles as the document's Title property. For headings, lists or code blocks, write Markdown and use Markdown to PDF.

A dropped file is accepted when its name ends in .txt or .text or your system labels it plain text, and the browse dialog is filtered to .txt and plain-text files; for anything else, paste the contents into the box or rename the file to .txt. Text in UTF-8 or UTF-16, with or without a byte-order mark, and older ANSI (Windows-1252) or Latin-1 text files are read with their accented letters intact, and a note under the box names the encoding when it was assumed rather than declared. Other legacy encodings, such as Cyrillic Windows-1251 or Shift-JIS, are not detected: they are read as Windows-1252 and come out with the wrong letters, so re-save those as UTF-8 first. Any character that still could not be read is counted in a warning under the box, because the PDF would print a replacement mark (�) in its place.

Not in one step: this tool converts one file or one pasted text at a time, and a multi-file drop loads just the first file. Either paste all the texts into the box one after another and convert once, or convert each file separately, reloading the page before the next one, then combine the results in order with Merge PDF.

Not fully. The page and its processing engine load over the network, and parts of the engine are fetched the first time a step needs them. The processing itself runs on your device. Drawing text in a script outside the Western European alphabet downloads a matching font from our site the first time it is needed, and that step needs a connection.

Up to 2 MB per file.

US Letter (8.5 × 11 in) portrait pages with 0.75 in margins and a page number (1 / N) at the foot of each page. There is no page-size or margin setting.