License & Data Attribution
ToPinyin is built on open data. This page lists every data source and its license, as required by the terms of use.
CC-CEDICT — Chinese-English dictionary
© MDBG and the CC-CEDICT contributors. Licensed under CC BY-SA 4.0. Source: mdbg.net. Dictionary entries on this site are derived from CC-CEDICT and share alike under the same license.
hanzi-writer-data — stroke order graphics
© chanind / hanzi-writer-data contributors, derived from the Arphic Public License fonts. Source: github.com/chanind/hanzi-writer-data (Arphic Public License).
Writing worksheets place these outlines on practice grids without changing the source paths.
Read the full, unmodified Arphic Public License.
The per-character source paths are available at /worksheet-data/{character}.json,
for example 学. The outlines are provided without warranty; review their suitability before teaching.
Make Me a Hanzi dictionary — character notes
Per-character definitions, decompositions, and phonetic/semantic component etymologies on the hanzi pages are derived from
skishore/makemeahanzi dictionary.txt (© its contributors),
licensed under the GNU LGPL v3.0 or later.
Heisig keywords — memory notes
The “How to remember it” keyword and composition notes are derived from rouseabout/heisig (MIT), based on Heisig & Richardson, Remembering Simplified Hanzi (© University of Hawai'i Press). The full story texts from the books are not reproduced.
pinyin-pro — pinyin conversion library
MIT License. Source: github.com/mos9527/pinyin-pro.
HSK character levels
HSK 3.0 (GF0025-2021) per-level character lists are transcribed from the community-maintained krmanik/HSK-3.0 compilation (CC BY-SA 4.0), which reflects the syllabus of the Chinese Proficiency Grading Standards for International Chinese Language Education (GF0025-2021, 教育部语合中心). Old HSK (2012) character grades are derived on this site from the official six-level word lists transcribed in glxxyz/hskhsk.com (MIT): a character is assigned to the lowest level whose official words contain it. HSK is a trademark of its respective right holders; this site is not affiliated with the test administrators.
Unihan Database — radicals & character variants
Kangxi radical classification and simplified/traditional variant forms on the hanzi pages are derived from the
Unicode Han Database (Unihan) and CJKRadicals.txt,
part of the Unicode Character Database. © Unicode®, Inc. Used under the
Unicode License & terms of use. Unicode and the Unicode Logo are
registered trademarks of Unicode, Inc. in the U.S. and other countries.
Pronunciation audio — synthesized at build time
Per-syllable pronunciation audio is generated at build time (scripts/gen-audio.mjs) with Microsoft Edge
neural text-to-speech voices (default: zh-CN-XiaoxiaoNeural) via the open-source
edge-tts (GPL-3.0) client. One audio file is synthesized per
pinyin syllable (about 1,200 files) by having the voice read a representative character for that syllable; files are
generated locally only when explicitly requested. Audio playback is available only for files included in the build;
missing audio is not replaced by browser speech synthesis, which may pronounce a different reading.
Generated audio requires pronunciation and service-terms review before publication. The client license does not grant rights to the service or its output.
Astro — site framework
MIT License. Source: astro.build.