data.placeholder.com/text

Placeholder text

Classic lorem ipsum, realistic filler copy in 17 languages, strings that break layouts on purpose, and pseudo-localization for your own UI text. All seeded: the same URL returns the same text forever.

Lorem ipsum

https://data.placeholder.com/text/lorem[?paragraphs=N | sentences=N | words=N | headlines=N][&format=][&start=][&seed=]

https://data.placeholder.com/text/lorem3 paragraphs, plain text, starting "Lorem ipsum dolor sit amet"
https://data.placeholder.com/text/lorem?paragraphs=5&format=html5 <p> elements
https://data.placeholder.com/text/lorem?words=20&start=random&seed=4220 words, not starting with "Lorem ipsum"
https://data.placeholder.com/text/lorem?sentences=3&format=jsonJSON with an items array and joined text
https://data.placeholder.com/text/lorem?headlines=4&format=markdown4 Markdown headings

Parameter	Values	Default
`paragraphs` / `sentences` / `words` / `headlines`	Pick one. Max 100 / 500 / 10,000 / 100. Same as `type=paragraph&count=N`.	3 paragraphs
`format`	`plain`, `html`, `markdown`, `json`	plain
`start`	`classic` (begins "Lorem ipsum dolor sit amet") or `random`	classic
`seed`	Any string up to 128 characters. Change it for different text.	`0`

Realistic copy in 17 languages

Neutral website and product copy that reads like the real thing, so you can judge typography, line length and layout in each locale. It is a parallel corpus: the same seed gives the same content in every language, so you can compare one layout across locales.

https://data.placeholder.com/text/<lang>[?type=paragraph|sentence|headline|product|ui][&count=N][&format=][&seed=]

Code	Language	Code	Language
`en`	English	`pt`	Portuguese (Brazil, `pt-BR`)
`sv`	Swedish	`pl`	Polish
`no`	Norwegian (Bokmål, `nb`)	`ar`	Arabic (right-to-left)
`da`	Danish	`he`	Hebrew (right-to-left)
`fi`	Finnish	`ja`	Japanese
`de`	German	`zh`	Chinese (Simplified, `zh-Hans`)
`nl`	Dutch	`ko`	Korean
`fr`	French	`it`	Italian
`es`	Spanish

Aliases: nb, pt-br, zh-hans, zh-cn, iw. Responses carry Content-Language, and HTML output is wrapped in <div lang="…" dir="…">.

https://data.placeholder.com/text/sv3 Swedish paragraphs
https://data.placeholder.com/text/de?type=headline&count=55 German headlines
https://data.placeholder.com/text/ja?type=product&format=jsonJapanese product names and descriptions
https://data.placeholder.com/text/ar?paragraphs=2&format=htmlArabic, with dir="rtl"
https://data.placeholder.com/text/fi?type=ui&count=1010 Finnish button and label strings

Types: paragraph (default 3), sentence (5), headline (5), product (3), ui (5); max 100 each, 500 for sentences. Shortcuts paragraphs=, sentences=, headlines= and products= work here too.

Stress-test strings

Hand-picked strings that break layouts, encodings and i18n: each with a note, a language tag, its length in code points, UTF-16 units and UTF-8 bytes, and the exact code points.

Kind	What it tests
`long-words`	Long compound words, unbroken strings, long URLs, e-mails and paths
`rtl`	Right-to-left scripts, mixed bidirectional text and bidi controls
`cjk`	Chinese, Japanese and Korean, full-width forms, astral ideographs
`emoji`	ZWJ sequences, skin tones, flags, keycaps, presentation selectors
`names`	Person names from 1 to 100+ characters: apostrophes, hyphens, particles, diacritics, scripts
`combining`	Combining marks, NFC vs NFD, stacked diacritics, Zalgo-style overflow, Indic and Thai
`whitespace`	Zero-width characters, no-break and fixed-width spaces, line separators, control characters
`numbers`	Locale number, currency, date and time formats, plus floating-point traps
`casing`	Case-mapping traps: Turkish I, sharp s, final sigma, digraphs, ligatures
`confusables`	Look-alike characters: homograph domains, styled letters, dashes and quotes
`scripts`	Language names in 30 scripts, for font fallback and line height
`pseudo`	Pseudo-localized UI strings

https://data.placeholder.com/text/stressevery kind, JSON
https://data.placeholder.com/text/stress/names?format=plainone name per line
https://data.placeholder.com/text/stress/rtl?count=5&seed=15 entries, shuffled by seed 1

format=plain puts one entry per line; some whitespace entries contain line breaks themselves, so prefer JSON for exact tests.

Pseudo-localization

Turn your own strings into accented look-alikes that are about 30 to 40 % longer and wrapped in brackets. You'll see at a glance which strings aren't translatable, which get cut off and which are glued together. Placeholders ({name}, {{count}}, %s, ${x}), HTML tags, entities, URLs and e-mail addresses are left intact.

https://data.placeholder.com/text/pseudo?text=Add%20to%20cart→ [ȦȦḓḓ ŧǿǿ ƈȧȧřŧ~]
https://data.placeholder.com/text/pseudo?text=Welcome%20back%2C%20%7Bname%7D!&format=json→ [Ẇḗḗŀƈǿǿḿḗḗ ƀȧȧƈķ, {name}!~] with lengths and expansion ratio

curl --data-binary @messages.txt "https://data.placeholder.com/text/pseudo"

?text= takes up to 4,000 characters; POST the raw UTF-8 text as the body for up to 64 KB (not cached). Options: expand=false, brackets=false, format=plain|json.

Notes

Everything is seeded and frozen (v1): same URL, same bytes, cached for a year.
The original copy is released as CC0; lorem ipsum is public domain.
/text returns a JSON index with every parameter, limit and language.
The old placeholder.com lorem ipsum tool at placeholder.com/text/lorem-ipsum redirects here.