About the flashcards
This edition contains 5,000 distinct Mandarin cards in simplified characters. Every card has an English study meaning, explicit pinyin, a word type, a Chinese example, the example’s pinyin, and an English translation.
Scope and selection
The collection begins with 300 original introductory cards and expands with 4,700 words and expressions adapted from licensed dictionary data. It covers everyday vocabulary, abstract ideas, idioms, and specialized meanings. It is not an official HSK syllabus or a definitive frequency ranking.
The 300 introductory cards were authored for this edition. The other meanings come from English Wiktionary via Kaikki/Wiktextract, with source-linked Wiktionary examples or Chinese-English Tatoeba examples. Words can have other meanings and grammatical roles; each card presents one study sense. Automated checks and editorial sampling exclude many mismatches, but cannot establish that all community examples are correct. The collection has not received an independent Mandarin-teacher review.
Topics and difficulty
The collection has 15 topics. Beginner, Intermediate and Advanced are suggested study groups. The original cards have authored assignments; sourced cards use meaning-based topic rules and a wordfreq selection signal. The labels are not exam levels or a proficiency assessment.
- Beginner: Start with useful everyday words, simple descriptions, and common conversation patterns.
- Intermediate: Expand everyday vocabulary and talk about experiences, interests, and a wider range of situations.
- Advanced: Explore less frequent words, idioms, specialist meanings, and more complex contexts.
Start here contains 150 cards, ordered from greetings and people through time, food, travel, and home. The 100 numbered sets contain 50 cards each. Four additional topic sets are views of the same vocabulary; they do not add duplicate cards.
Pinyin and pronunciation
Every word and example has explicit pinyin. External readings come from Wiktionary or reviewed Tatoeba transcriptions; missing readings are never filled with machine transliteration. Unmarked syllables have a neutral tone. Sources differ in word spacing, neutral tones and whether tone changes for 一 and 不 are shown; third-tone changes are usually left in dictionary form, such as nǐ hǎo. Spoken delivery can vary with context.
Search accepts characters, English, and pinyin with tone marks, numbers, or no tones: nǐ hǎo, ni3 hao3, and nihao find the same greeting. For ü, you can also type u: or v. This permissive search behavior does not set the grading rules in course exercises.
Word and sentence buttons request generated Mandarin speech from the site’s speech service when configured, with a device voice as a fallback. Voice availability depends on the service and your browser. Pinyin is a reading aid; listen to whole phrases as well as individual words.
Practice and progress
Practice includes all cards matching your filters and loops back to the beginning. You can reverse the direction, shuffle, and mark words familiar or unfamiliar. Familiarity is shared across sets and saved in this browser. The browse view shows 20 cards per page. There are no learner accounts or automatic proficiency scores.
Sources and licenses
Kaikki’s extraction of English Wiktionary supplies 4,700 meanings and 3,227 sense-linked usage examples. The dictionary dump is dated October 1, 2026, extracted October 25. These adaptations are attributed to Wiktionary contributors and shared under CC BY-SA 4.0. Each card links to the source page and its contributor history. Source quotations were excluded.
Tatoeba contributors supply 1,473 bilingual examples under CC BY 2.0 France. Each example links to both sentence pages for contributor attribution. Pinyin comes from a transcription with a named reviewer; this does not guarantee that every reading is correct.
Traditional forms are converted to simplified with OpenCC; numbered pinyin is converted to tone marks. The source repository preserves original strings, sense identifiers, sentence IDs, transcription reviewers and input hashes. Dictionary cross-references are shortened for display. Regional wording and pinyin conventions may remain.
wordfreq by Robyn Speer helps prioritize the sourced vocabulary; its data and derived selection signals are CC BY-SA 4.0. Full upstream attribution is available. This signal does not prove a word’s exact frequency or difficulty.
Data and provenance
Download the complete Mandarin study data (JSON). The export includes pinyin, bilingual examples, topics, and stable identifiers. The source repository contains the authored input, selected licensed source records, a reproducible offline generator, source hashes and validation reports.