← Back to Star Japanese

Vocabulary sources

The course contains 8,450 distinct cards and covers 7,811 N5–N1 entries from OpenJLPT 0.3.0. You can search for alternate spellings of a word. Words that sound the same but have different meanings have separate cards.

Reference entries and course cards by level
LevelReference entriesReference cards in filterCourse additionsCards in filter
N56746743677
N463062956685
N31,6591,6482111,859
N21,7781,7593782,137
N13,0703,065443,109

The 3,070 N1 entries correspond to 3,065 cards. Three batches added 901 cards each. Of the 362 existing cards, 345 moved from the N2 supplement to N1. The other 17 also appear at lower levels and share favorites and review history across both filters.

The course includes 7,758 reference cards and 692 additional cards. Since the 2010 revision, the JLPT has not published a word-by-word exam list.Textbook indexOfficial JLPT FAQ

Additional words in current use

These 44 additions span 11 categories, including teamwork and supply and demand. Twenty appear in the 2012 or 2018 official N1 workbooks, another 23 in the revised TRY! N1, and ずさん in N1単語3000. The map marks them as N1 additions.

To check current formal usage, we searched the Japanese Diet proceedings from 2022 to 2025 for each word and its inflected forms.Guide to the Diet proceedings search

View proceedings for the 44 words

Scenes for common words

Common words include our N5 and N4 words, the 44 additions above and higher-level words matching a top-5,000 entry in BCCWJ written Japanese or CEJC everyday speech by form, reading and part of speech. Everyday expressions from shopping, travel, household, school and work examples are also included. Each picture follows a specific sense and example. Actions loop when movement conveys the change. Illustrations show objects, states and tone.NINJAL BCCWJ short-unit word list · CEJC short-unit word list (September 2022)

For corpus design and collection, see Hanae Koiso et al., “「日本語日常会話コーパス」の設計と構築,” NINJAL Research Papers 24 (2023), pp. 153–168.

Levels, meanings and examples

Levels and readings follow OpenJLPT, drawing on Jonathan Waller’s JLPT Resources and EDRDG’s JMdict. The course revises variant spellings, parts of speech, Chinese glosses and examples with highlights. Level references and derived data are available under CC BY-SA 4.0.

The multiple-meaning collection covers 2,722 cards and 7,660 common senses with Japanese–Chinese examples. Each sense has a separate example with furigana and a translation. Readings are distinguished by usage and corrections are made when older examples do not match the card reading. Select a sense to see its form, reading and translation. All senses share audio, favorites and study history. The presentation of words in different contexts was inspired by Kaka’s N2 vocabulary video。

Examples by sense and batch
BatchCards revisedNew senses (one example each)
Batch 13861,115
Batch 2115205
Batch 3267667
Batch 4146284
Batch 5334656
Batch 675102
Batch 73038
Batch 8356492
Batch 9395555
Batch 10376540
Batch 1196120
Batch 123134
Batch 13115130

Common polysemous words were selected using NINJAL written word frequencies and checked against modern senses in JMdict. Examples reflect everyday life and study. Frequency describes the word as a whole.BCCWJ frequency list

Chinese glosses were checked against Japanese-Chinese-thesaurus (Unlicense) and edited for each sense. Examples include original Japanese–Chinese sentence pairs and contributions from Tatoeba. Furigana was added for each example; some Chinese translations were revised or converted to simplified characters.

Course translations

English, Korean and Vietnamese meanings and example translations are study aids. The course uses offline-generated translations with spot checks and existing human translations when a matching source is available. Refer to the Japanese example for usage in context.

Scene illustrations and Japanese audio

This edition has offline scenes for 13,388 senses across 8,450 cards. Changing the sense also changes the example and scene. Actions or changes that convey meaning use loops; objects, states or relationships appear as illustrations. Some props use Lucide SVG icons, with credits and license in word-stage.js. Japanese word audio was generated with VOICEVOX Nemo for offline playback.