Dictionaries — entry 010 of 12
Wiktionary
Wiktionary exposes its full multilingual dictionary — definitions, etymologies, pronunciations, and translations spanning thousands of languages — through the same MediaWiki Action API that powers Wikipedia, with no key required for read access. Cross-origin browser requests need the documented origin=* query parameter rather than a bare Origin header, since that's what actually unlocks the API's CORS headers.
Wiktionary exposes its full multilingual dictionary through the same MediaWiki Action API that powers Wikipedia — definitions, etymologies, and pronunciations for any entry, with no key required for read access.
GreatAPIs Score
Embed this badge
<a href="https://greatapis.com/api/wiktionary/"><img src="https://greatapis.com/badge/wiktionary.svg" alt="Scored 94 on greatapis.com"></a>Auth quickstart
- No key needed — call it now.
Your key is stored only in this browser (localStorage) and sent directly to the API — never to greatapis.
Get a plain-text extract for a word
GEThttps://en.wiktionary.org/w/api.php?action=query&titles=hello&prop=extracts&explaintext=1&exchars=300&format=json&origin=*
{
"batchcomplete": "",
"query": {
"pages": {
"4803": {
"pageid": 4803,
"ns": 0,
"title": "hello",
"extract": "\n== English ==\n\n\n=== Alternative forms ===\nhullo (UK)\nhallo, hollo (UK, dated)\nhilloa (obsolete)\n\n\n=== Etymology ===\nFirst attested in 1826. From holla, hollo (attested 1588)..."
}
}
}
}Try it
Developer reference
https://en.wiktionary.org/wWikimedia guidance: unauthenticated clients should stay below 5 requests/second (concurrency 1), authenticated clients up to 10 requests/second (concurrency 3)
- GET/api.php?action=query&titles={title}&format=json
Gotchas & limits
- Browser/cross-origin requests need
origin=*in the query string — a bareOriginheader is not enough to unlock CORS prop=extractsreturns every language section on the page (English, French, Fula, ...) unless you constrain it —excharsorexsentencestrims the response to a usable size- Pages are keyed by numeric
pageid, not by the title you requested — look the entry up by iteratingquery.pages, not by indexing on the word - No API key exists to request — anonymous traffic is rate-limited by IP instead