Open data · ~50,000 knowledge objects
Matrix Hanzi organizes ~50,000 structured knowledge objects into 31 sets: HSK 1–6 vocabulary (Simplified), TOCFL 1–7 vocabulary with Star extensions (Traditional), 26,000+ sentence patterns and 3,008 idioms. Every object carries Hanzi, pinyin, meaning, structure analysis, mnemonics, compound words and examples — plus exclusive line-art illustrations for most entries.
Pick a set below to see its role in your learning journey, preview 5 real data objects, then open it directly in the embedded learning modes.
The standard vocabulary track for HSK 3.0 learners — the foundation for study, work and certification in mainland China.
For learners on the Taiwan track. Star (★) sets are supplementary vocabulary extending each band.
Learn words in real context: each pattern includes structure analysis, usage notes, key phrases and similar sentences.
Traditional-script context for the Taiwan track, from basic conversation to sophisticated expression.
The full idiom treasury with literal and figurative meanings, mnemonics, classical allusions and example sentences.
Pick a dataset once on the home page; the same session carries across every mode.
Yes. All HSK/TOCFL vocabulary, idioms, sentence patterns, image flashcards, typing practice and games are free with no login. Only deep AI pronunciation analysis is limited, because it has real API costs.
Each dataset works across 6 modes: Learn & Analyze, image Flashcards, Active Recall (Pro Mode), Reflex Game, Hanzi Typing and Speaking practice. Pick a set once on the home page and keep it across modes.
HSK uses Simplified Chinese and fits study, work and exams for mainland China — it is the most common system. TOCFL uses Traditional Chinese and fits Taiwan. Matrix Hanzi covers both, including the Star extension sets for TOCFL.