Hanzi Sheets

About & licenses

Hanzi Sheets is a free, no-sign-up worksheet generator for families and classrooms teaching children written Chinese. It runs entirely in your browser — your text and mistake book never leave your device.

Reading, glyph and stroke coverage

The 11,518-word reference uses Mainland Mandarin dictionary readings, including lexical neutral tones. Traditional word forms and polyphones resolve at word level; unlisted text uses character fallback, so names, rare words and ambiguous single characters need a teacher’s check.

Script and accent are independent. The optional Taiwan variants setting covers 9 supported word variants: 爸爸、謝謝、頭髮、們、我們、你們、他們、她們、它們. Other words retain the Mainland reference. For example, Mainland 頭髮 is tóu fa, while Taiwan is tóu fǎ. These use four checked dictionary records plus the personal-pronoun plural-ending rule. Facts were checked against MOE’s 爸爸, 謝謝, 們, and 頭髮 records; no prose, audio or animations copied.

There are 6,373 simplified and 3,805 traditional stroke files in this release. Make Me a Hanzi follows PRC stroke order, including for traditional characters; these are not Taiwan MOE-standard diagrams. Missing diagrams are omitted, never guessed. Taiwan MOE’s online animations carry a noncommercial license, so they are not reused here.

The 1999 Arphic regular Kai fonts cover 6,312 of 6,499 requested simplified characters and 6,409 of 6,597 traditional characters. Of the missing font characters, 61 simplified and 21 traditional characters have licensed stroke outlines used as tracing models; for the remaining 126 simplified and 167 traditional characters the cell says “no glyph”. Unlisted characters use the same outline-or-explicit-placeholder policy. We do not silently trace a system fallback font.

These are regular Kai fonts rather than cursive handwriting. Neither font is certified against every current school textbook or Taiwan MOE glyph standard. Check the model for the curriculum you teach. Exact font coverage and source checksums.

Data & font attribution

AssetSourceLicense
Simplified stroke-order data hanzi-writer-data (Hanzi Writer project) Arphic Public License
Traditional stroke-order data Make Me a Hanzi (skishore/makemeahanzi) Arphic Public License / LGPL
Simplified ↔ Traditional mapping OpenCC (STCharacters / TWVariants) Apache-2.0
HSK 2.0 / 3.0 vocabulary complete-hsk-vocabulary MIT
English dictionary definitions (within HSK vocabulary) CC-CEDICT (via MDBG, used by complete-hsk-vocabulary) CC BY-SA 3.0
YCT 1–4 vocabulary YCT outlines (汉考国际 / Chinese Testing International), mirrored by public word lists Public exam syllabus content
Pinyin & zhuyin annotations Word readings reconciled with the CEDICT-derived vocabulary; pypinyin character fallback. Selected Taiwan word variants checked against MOE. MIT
AR PL KaitiM GB (regular Kai, Simplified) Arphic Technology (1999); WOFF2 family HSKaiGB Arphic Public License 1999
AR PL KaitiM Big5 (regular Kai, Traditional) Arphic Technology (1999); WOFF2 family HSKaiBig5 Arphic Public License 1999
Common-character table (通用规范汉字表) Chinese national standard character tables (2013 / 1988) Public domain

Full license texts

The family relationship calculator uses a local, unmodified copy of mumuy/relationship 1.2.9, commit 35c44427befb6fd59e16f583b5b92a680a7948db, by Haole Zheng (MIT). Its separate 52-entry vocabulary selects pinyin from mozillazg/phrase-pinyin-data at cee0ed6e6e4898580cafd2bd5e3723e20b214aa0 (upstream MIT notice) and Jyutping from Rime Cantonese by CanCLID at 259f0e48bba840c3a2e0d117539e96937f3d89bc (CC BY 4.0). Modified by selecting readings, pairing character forms, composing explicitly marked readings, and adding relationship explanations. No endorsement is implied. Full modification and source-lineage notes and individual source records accompany the data. No recordings are included; independent human language review is pending.

Privacy

Everything is generated client-side. The mistake book and the daily free quota are stored only in your browser’s local storage; no account, no tracking of worksheet content.