About
Where the data comes from
Jiyomi’s dictionary is built from two files of the Electronic Dictionary Research and Development Group:
- JMdict, the dictionary of common words, licence CC BY-SA 4.0.
- JMnedict, proper names (people, places, companies, works), licence CC BY-SA 3.0, © Jim Breen & the EDRDG.
Also: stroke order from KanjiVG (Ulrich Apel, CC BY-SA 3.0), kanji data from KANJIDIC2 (EDRDG, CC BY-SA 4.0), and word frequencies from wordfreq (Robyn Speer, CC BY-SA 4.0, built from Wikipedia, OpenSubtitles, the web, Twitter, Reddit), which order search results. Names of works, characters and places that these dictionaries lack come from the titles of the Japanese Wikipedia (Wikipedia contributors, CC BY-SA 4.0), and a name’s entry shows, from your browser, the start of its article.
The database published by Jiyomi is a derivative work: it is distributed under the same licences, CC BY-SA, each that of its source. The English meanings are JMdict’s own, written by its contributors.
Download the database
The SQLite file the app and this site use is public: github.com/AStoicBoy/jiyomi-dico (manifest.json describes each language).
Also in French
Jiyomi’s French dictionary is the largest freely available: JMdict carries French on 15,336 entries; Jiyomi wrote the 203,465 others, directly from the Japanese. jiyomi.app, in French.