About Waffuru
Waffuru was built for learners who want to master Japanese faster and more effectively. Traditional flashcards often waste your time on obscure words — we take a smarter approach.
By combining word frequency analysis and science-backed spaced repetition, Waffuru ensures you learn the right words at the right time.
Whether you're preparing for JLPT, studying abroad, or simply learning for fun, Waffuru helps you stay motivated and make steady progress every day.
Data Sources
The data used in Waffuru comes from a variety of open source projects. We are immensely thankful to the people who work on these projects and make them available for everyone studying Japanese to benefit from.
JMdict
JMdict, created by Jim Breen and maintained by the Electronic Dictionary Research and Development Group (EDRDG), is a comprehensive Japanese-English dictionary with hundreds of thousands of entries. It is the source of the word data in Waffuru.
KANJIDIC
KANJIDIC, also from Jim Breen and the EDRDG, is a database of over 13,000 kanji that includes readings, meanings, stroke counts, and extensive metadata. This is the source of kanji information in Waffuru.
RADKFILE / KRADFILE
RADKFILE/KRADFILE, maintained by Jim Breen and the EDRDG, provides a decomposition of kanji into their visual components and radicals. This data powers the radical information displayed on kanji cards.
KanjiVG
Stroke order data comes from the excellent KanjiVG project by Ulrich Apel. KanjiVG provides vector graphics with stroke order, direction, and component information for kanji characters, released under a Creative Commons Attribution-Share Alike 3.0 license.
Jonathan Waller's JLPT Resources
Information about which words and kanji belong to which JLPT level comes from Jonathan Waller's JLPT Resources, which provides vocabulary and kanji lists organized by JLPT level (N1–N5).
FSRS
Waffuru's spaced repetition system is powered by the Free Spaced Repetition Scheduler (FSRS) algorithm, developed by the Open Spaced Repetition community. FSRS is a research-backed algorithm that optimizes review timing to maximize retention with minimal study time. We use ts-fsrs, the TypeScript implementation, released under the MIT license.