What this tool is for
Use it for fast lexical lookup. Each supported entry is backed by a bounded tool corpus sourced from external language datasets rather than the repo's JLPT study lists.
Search supported Japanese words by kanji, kana, romaji, or English meaning.
Use it for fast lexical lookup. Each supported entry is backed by a bounded tool corpus sourced from external language datasets rather than the repo's JLPT study lists.
Search for a word, compare the readings and meanings in the result cards, and then move into JLPT study from the matching level tag when it is available.
The long-standing community favourites are Jisho, Tangorin, JapanDict, Takoboto, and Jim Breen's WWWJDIC — all built on the same JMdict source data. This GyanMirai dictionary covers the same core JMdict lookups but layers JLPT level tags on every entry and shows Tatoeba example sentences inline, which makes it faster for learners who are studying toward a specific level. For rare, archaic, or highly technical vocabulary, Jisho still has broader coverage; for day-to-day JLPT learning the GyanMirai dictionary is the fastest path from query to study page.
You can search by kanji, hiragana, katakana, romaji, or English meaning in the same input box. Paste a kanji you saw in the wild and the dictionary returns every matching entry with its reading. If you cannot type the kanji, use the Kanji Lookup by Radical tool linked below to find it first, then paste the result here. Romaji search is supported for anyone who has not learned kana yet — typing "tabemasu" finds 食べます correctly.
Most common entries have at least one Japanese + English example sentence drawn from the Tatoeba project. Rare or highly specialised entries may not have sentences yet; the dictionary shows this state explicitly rather than pretending coverage. If an entry is missing an example you need, searching for a closely related word usually surfaces a sentence that uses the target in context.
Every entry is tagged with the JLPT level (N5 through N1) at which the word typically appears, using the widely accepted N-level vocabulary lists. This lets you filter for study-level-appropriate vocabulary: if you are preparing for N3, you can scan the results and focus on the N3 and below entries first. Words that do not appear on any JLPT list are marked as uncategorised.
Roughly 3 000 words gives you working fluency for everyday conversation, matching the vocabulary load of JLPT N2 and most of N3. That number also tracks with the Pareto-style claim that the 3 000 most-used Japanese words cover 85–90% of casual written and spoken language. For reading news, novels, and passing JLPT N1, you need closer to 10 000 words. The dictionary plus the JLPT vocabulary study pages linked below are built to walk you through that curve one level at a time.
The UI is fully responsive and static, but word lookups go through a server endpoint that returns the JMdict entries and Tatoeba examples. You need an internet connection for the search itself. Once you have the results on screen, you can read, copy, and navigate to study pages without further network calls. For a truly offline Japanese dictionary, dedicated desktop or mobile apps (Takoboto on Android, Imiwa on iOS) bundle the same dataset locally.
Looking up Japanese because you are planning a Japan move? Start with the application path that fits your level.
The lab-first admissions process — and why it differs from US/UK applications.
Read the guide →Step-by-step MEXT 2027 application: timeline, documents, embassy vs. university track.
Read the guide →Current list of English-track programs at top Japanese universities.
Read the guide →