About

Why we built it

Most JLPT tools were designed for English speakers. Grammar is explained in English, examples translate through English, and there is nothing tailored to the real pain points Chinese speakers hit — Japanese-Chinese false friends, on/kun reading patterns leveraging Chinese phonetic intuition, or the specific kanji differences between 日本漢字, 繁體, and 简体.

Tadorimichi fills that gap. Every vocabulary entry is annotated by Claude Opus with learner notes written for Chinese speakers. Every kanji card foregrounds readings and phonetic groups. Every grammar point is cross-referenced with similar patterns so you learn the nuance.

What we believe

Passing JLPT is the entry point, not the finish line. We would rather help you read a novel, write a business email, or watch a drama unsubbed than gate you behind artificial progress barriers. What we ship today is the full core experience — vocabulary, kanji, grammar, reader, mock exam — with progress saved locally, no signup required. The product is in beta and the business model will be announced separately.

Data sources

The 8,047-word vocabulary core, 2,682-character kanji core, 723 grammar points, and all generated articles are originally compiled by Tadorimichi. For words outside the core set we fall back to JMdict and KANJIDIC2 (EDRDG, CC BY-SA 4.0). See Attribution for the full breakdown and license terms. We are grateful to the volunteers who built and maintain those resources.

Who we are

Tadorimichi is operated by Mason AI Lab. For questions, feedback, or partnership inquiries, write to us at [email protected].