Home/Open data
The same curated data that powers the LinZi app and this site - vocabulary lists, cross-strait pairs, measure words and the story catalog - as free CSV and JSON downloads under CC-BY-SA. Every file ships a sha256 in the manifest.
All 11,036 words tagged in the HSK 3.0 syllabus - levels 1-6 plus the combined 7-9 advanced band - each with pinyin, zhuyin and curated English definitions. Sorted by level, core-10,000 vocabulary first.
All 7,187 words in Taiwan's TOCFL 八千词 syllabus, Traditional-first. The tocfl_level column encodes the two Novice bands as 0.1/0.2 and Levels 1-5 as integers; a tocfl_band column spells out the label. With pinyin, zhuyin and English definitions.
953 words Taiwan and the Mainland say differently - 计程车 vs 出租车, 软体 vs 软件 - as verbatim rows: Taiwan term, MOE alternates, Mainland term, both scripts, English gloss, plus a stable slug.
100 curated "don't confuse these" pairs - look-alike characters (未/末, 己/已) and easily-mixed words (还是/或者, 二/两) - each with a one-line teaching note. The same list the /compare/ pages and the app share.
133 classifiers - 个, 张, 条, 把 and the rest - with the dictionary nouns each one takes, easiest first. Pulled from the same classifier index that drives the app's measure-word trainer.
Metadata for all 1,760 graded stories in the corpus: title in both scripts, English title, category, level 1-6, the free/published flag, word and character counts, and the canonical URL.
Everything in /data/files/ is CC-BY-SA
4.0: free to reuse and remix, commercially too, as long as you
credit LinZi and keep the same license. Drop this where you use
the data:
Vocabulary data from <a href="https://www.linzi.app/data/">LinZi</a> - linzi.app/data/Plain text: Vocabulary data from LinZi - linzi.app/data/ · licensing details and per-file checksums: LICENSE.txt · index.json
Curated lists and compilations are ours - CC-BY-SA 4.0, attribution "LinZi (linzi.app)". English definitions derive from CC-CEDICT (CC-BY-SA), which is what puts every file under share-alike terms. The underlying word lists are government syllabi and freely redistributable: HSK 3.0 (《国际中文教育中文水平等级标准》 GF0025-2021, PRC Ministry of Education) and TOCFL 华语八千词 (SC-TOP / Taiwan Ministry of Education). The story catalog is metadata only - text sources are credited on the credits page.
These aren't scraped lists - they're generated at site-build time from the same dataset that ships inside the app: search them live in the dictionary, compare the easily-confused pairs, drill the measure words, browse Taiwan vs Mainland vocabulary or read the graded stories. TOCFL entries map to the band lists and HSK entries to the level lists.