Nihongo is built on open Japanese-language datasets. All of the learning content here is derived from the projects below — with thanks to the authors and communities who share them under open licenses.
Vocabulary entries (kanji, kana, part of speech, English glosses) and kanji data (on/kun readings, meanings, stroke counts). Property of the Electronic Dictionary Research and Development Group (EDRDG) and used under the Group's licence. We consume the data via the jmdict-simplified JSON distribution.
Per-kanji SVG stroke paths and stroke order, used for the writing animations. © Ulrich Apel.
Japanese example sentences with translations. Each sentence's ID is retained in our content files for per-sentence attribution.
Assignment of vocabulary, kanji, and grammar to levels N5–N3. Compiled by Jonathan Waller (tanos.co.uk); we also reference the GitHub mirror (elzup/jlpt-word-list).
The kanji assigned to level N3 — the literals only; readings, meanings, and stroke counts come from KANJIDIC2. Compiled by mund-tandem.com (now offline) and distributed with the tanos.co.uk JLPT materials, which publish no N3 kanji list of their own.
The JLPT level lists are community-sourced and unofficial. Since the 2010 reform, the Japan Foundation no longer publishes content specifications for each level, so the N5–N3 assignments here are community estimates and may contain errors.
Full attribution file (English): ATTRIBUTION.md