Licenses & credits

CC-CEDICT — CC BY-SA 4.0

Community Chinese–English dictionary — readings, meanings and traditional/simplified pairs.

Make Me a Hanzi — LGPL / Arphic PL

Stroke-order graphics, radical decomposition and character etymology.

Hanzi Writer — MIT

The stroke-by-stroke writing and animation engine used in the writing trainer.

HanziLookupJS — MIT

Handwriting recognition that turns your drawn strokes into a character, fully offline. Its recognition data (mmah.json) is derived from Make Me a Hanzi stroke data and carries that project's LGPL / Arphic PL terms, not the MIT of the engine itself.

Unicode Unihan — Unicode License

The Unicode Han database — canonical readings and character properties.

中華民國教育部（Ministry of Education, R.O.C.）《國語辭典簡編本》 — CC BY-ND 3.0 TW

Audio for characters and words. Recordings by Taiwan's Ministry of Education, data version 20260626. Files are served unmodified; the app only limits playback to the headword.

使用說明 — the official usage notice, retained in full as the licence requires

OpenCC — Apache-2.0

The STCharacters table, used to convert simplified characters into traditional ones when you paste in text. Every substitution it makes is listed back to you.

Roboto — SIL OFL 1.1

The interface face, used for Latin text and for pinyin. Self-hosted rather than loaded from Google, so no request leaves for a third party when you open a page.

Noto Sans TC — SIL OFL 1.1

The default face for traditional characters. Also self-hosted.

全字庫正楷體 · TW-Kai — SIL OFL 1.1 · 政府資料開放授權條款 v1

Taiwan's own kaishu brush face from the CNS11643 全字庫, published by 數位發展部. This is the face behind the handwriting option, and its 臺 outline is what the site logo is drawn from. We ship a subset of it: the file is reduced to the characters this app actually uses, and the logo embeds outlines taken from it, so the copy here is modified rather than original.

Ministry of Education standard character forms, the 4808 common-character set and character frequency.

The official TOCFL graded vocabulary lists, level by level.

National Academy for Educational Research — the TBCL benchmark, COCT corpus and general word-frequency table.

Zhuyin symbol recordings from the MOE 國語注音符號手冊 open-components bundle, released separately under CC BY 4.0. 2017 © 教育部,國語注音符號手冊-開放部件.

The normative list of stroke names behind the basic and compound stroke inventory.

Rainfall scales, typhoon categories and the UV index bands.

Tea classification, processing stages and oxidation levels for Taiwanese cultivars.

The lunar calendar and solar term tables the festival dates are computed from.

Traced ancient character forms — oracle bone, bronze and seal script. Released into the public domain.