Lorem ipsum generator

This lorem ipsum generator returns the same text for the same seed

Most placeholder generators hand you fresh random text on every press. That is fine until two people look at the same mockup. A designer exports a frame, a developer rebuilds it from the same settings, and the two screens now hold different sentences of different lengths — so the only visible difference between the versions is the one nobody meant to introduce.

Here the output is a function of the seed. A seed is a short word you can say out loud in a review, and with the same seed, the same unit and the same count, this page returns byte-identical text every time, in any browser. Changing the count and changing it back is not a reroll. New text comes from a new seed, which is what the New seed button is for, and the seed that produced a result is printed on the receipt beside it so a layout can be reconstructed months later. If you ever want to confirm that two runs really did match, paste both into the text compare tool and look for zero differences rather than trusting your eye on nine paragraphs of Latin.

Where these words actually come from

The vocabulary is not a copied word list. It is derived, in code, from three named passages that ship with this page: sections 32 and 33 of book one of Cicero's De finibus bonorum et malorum, a treatise on the ends of good and evil written around 45 BC, and the garbled typesetting passage that begins Lorem ipsum dolor sit amet. Deduplicating and sorting the words of those three passages produces a pool of 210 distinct words, and a word can appear in the output only if it appears in one of the passages. That is a property of the program rather than a promise in this paragraph, and a test fails the build if it stops being true.

Cicero has been dead for more than two thousand years, so the Latin is free of copyright by any measure, and the passages are reproduced here rather than fetched from anywhere. Working from a named source also makes the pool inspectable: run a sample through the word frequency counter and you can see exactly how often a 210-word vocabulary repeats itself at the length you asked for.

Lorem ipsum is a corruption, not a quotation

This matters if you have ever wondered what the text means. Cicero's section 32 contains the phrase qui dolorem ipsum, quia dolor sit, amet — roughly, "who loves pain itself, because it is pain". Typesetters cut the first syllable off dolorem, dropped and respaced words, and turned adipisci velit into adipiscing elit and eius modi tempora incidunt into eiusmod tempor incididunt. What survived is the familiar passage — and lorem is not a Latin word at all.

That is the whole point of the material. Placeholder text is supposed to be unreadable enough that a reviewer looks at the type, the measure and the rag instead of the argument. Real Latin would still be parsed by anyone who reads Latin; a corruption of it is safely inert. The route the corruption took into the trade is usually told as dry-transfer lettering sheets in the 1960s and page-layout software after that. This page does not depend on that story being exact, because both passages are here to compare directly: the classic opening is a checkbox, not a hardcoded default, and switching it off gives you text drawn only from the pool.

Four units, and the count is exact

Paragraphs, sentences, words and list items are four different questions, so they are four modes rather than one mode with a multiplier. Ask for 40 words and you get 40 words, not "about 40". Ask for 12 sentences and you get 12 sentences ending in 12 periods. Paragraphs hold between three and seven sentences each, and list items are short capitalized fragments without terminal punctuation, because that is how list items are normally written.

There is no upper limit on the count. The page estimates how many characters the current settings will produce. Ordinary amounts appear as you type; a run large enough to be worth thinking about waits for Generate text and says why first, since a browser tab runs out of memory long before it runs out of time. Nothing is refused for being big. When you want the precise measurement rather than the estimate, the receipt gives word and character totals for what you actually got, and the character counter will break the same text down into graphemes, code points and UTF-8 bytes if a field limit is defined in one of those.

HTML mode builds structure, not a string with tags in it

Choosing HTML changes the shape of the result, not just its wrapper. List items become a ul with one li per item. Paragraphs become sibling p elements. Sentences and words become a single p, because five sentences are one block of prose and emitting five paragraphs there would answer a question you did not ask. The markup is parsed as markup, in both Chromium and WebKit, before it ships.

The output is deliberately unindented so it pastes cleanly into a template. If you would rather have it laid out, the HTML formatter will indent it, and if you have inherited a template already full of placeholder markup and need the words without the tags, HTML to text strips them while keeping the block structure readable. To fold multi-paragraph output onto a single line for a JSON fixture or a spreadsheet cell, remove line breaks does that without collapsing the words together.

Nothing leaves this tab

Generation is a few hundred bytes of arithmetic over a word list that is already in the page. There is no request carrying your settings, no account, no history of what you generated and no server that could keep one. Copy puts the complete text on your clipboard even when the preview on screen is clipped, and download builds the file from the same string in the browser. Reloading the page clears everything except what you chose to keep.