Chinese character frequency lookup. The 200 most-used characters with rank and cumulative coverage, counted by RECATOOLS from Chinese Wikipedia (1 September 2026 dump, CC BY-SA 4.0). Runs in your browser.
Chinese Character Frequency & Coverage (汉字使用频率)
How to use
View the Top 50
See the 50 most frequent characters at a glance — in Chinese Wikipedia these 50 make up about a quarter (24.9%) of all characters.
Search by character
Enter a Hanzi, pinyin, or English meaning — tool returns rank and cumulative coverage %.
Gauge text difficulty
When writing, check the rank distribution of characters used — more high-frequency = more readable.
Prioritise high-frequency learning
Beginners should master the top 200 first — fastest path to reading competence.
Character Frequency: Why the Top 1,000 Characters Matter Disproportionately
Chinese character frequency is extremely uneven: a small handful of characters appear constantly, while most almost never do. Counting about 776 million characters of Chinese Wikipedia: the top 100 characters make up 35.8% of the text, the top 1,000 make up 88.9%, and the top 3,500 make up 99.5%. Knowing the first 3,500, you meet on average fewer than one unfamiliar character in a hundred.
The learning leverage this creates
For learners, that is the case for learning characters in frequency order: the first 500 already cover 73.6% of Wikipedia text, and the next 3,000 raise that only to 99.5%. When the 1988 List of Frequently Used Characters in Modern Chinese was tested on a two-million-character sample, its 2,500 common characters covered 97.97% and the next 1,000 added 1.51%.
Corpora and frequency statistics
This tool\'s figures are RECATOOLS\' own count: every article in the 1 September 2026 dump of Chinese Wikipedia (about 1.55 million, redirects excluded), with templates, tables and reference sections removed and traditional characters converted to simplified with OpenCC before counting. Wikipedia text is licensed CC BY-SA 4.0, and this list is offered under the same licence, with credit to the contributors of Chinese Wikipedia.
Note: the corpus decides the ranking. Wikipedia is encyclopedic, so 年 (year), 国 (country), 区 (district) and 市 (city) rank high while 我 (I) ranks only 465th; in corpora built from fiction and conversation, 我 is usually in the top ten. For spoken Chinese, treat this list as a guide to written Chinese.
In Chinese Wikipedia, one character in thirty-seven is 的
的 is the most frequent character, 2.72% of all characters in Chinese Wikipedia — about one in 37. In corpora made of fiction and conversation its share is higher.
The top 10 characters make up 10.4% of Chinese Wikipedia: 的, 年, 为, 在, 一, 国, 中, 人, 是, 大. 年 (year) is second because encyclopedia articles are full of dates.
Frequency falls steeply, but less neatly than Zipf's 1/N: in our count the 10th character, 大, appears about a fifth as often as the first, 的 — not a tenth.
The top 1,000 characters make up 88.9% of Chinese Wikipedia, and the top 3,500 make up 99.5%. Past the first 1,000, each further character adds less than a tenth of a percentage point.
The 1988 List of Frequently Used Characters in Modern Chinese has 2,500 common and 1,000 less-common characters. Tested on two million characters of 1987 newspapers and magazines, the common ones covered 97.97% and the rest added 1.51% — 99.48% in all.
Total Chinese character count far exceeds 3,500. The Kangxi Dictionary lists 47,035. The Hanyu Da Zidian lists 55,000. Unicode CJK encodes 97,000+. But the vast majority are historical, name-specific, or technical characters rarely used in modern daily writing.
The same character has different frequency in different corpora. News articles → 政, 济, 府 rank higher. Classical literature → 之, 乎, 也. Online chat → 啊, 呢, 哈. This tool uses general-purpose modern corpus data.
Simplified vs. traditional doesn't affect frequency rankings (1:1 mapping), but stroke count differs dramatically. 龍 → 龙 (16→5 strokes), 邊 → 边 (19→5). This was the original motivation for simplification — making the most frequent characters easiest to write.
Recognising ≠ writing. An educated Chinese reader can recognise 4,000-6,000 characters but may only be able to write 2,500-3,500 unaided. "IME dependence" — losing handwriting fluency due to typing-only use — is a real modern phenomenon.
Pairs with RT-CHN-031 (Chengyu Dictionary) and RT-CHN-032 (HSK Vocabulary) — the three pillars of systematic Chinese learning.
Frequently Asked Questions
-
RECATOOLS counted it: every article in the 1 September 2026 dump of Chinese Wikipedia — about 1.55 million articles and 776 million characters — with templates and references removed and text converted to simplified with OpenCC. The data is licensed CC BY-SA 4.0.
-
The corpus decides the ranking. Wikipedia is encyclopedic: 年 (year) ranks 2nd and 月 (month) 12th, while 我 (I) is 465th and 你 (you) 969th. In corpora built from fiction and conversation, 我 is usually in the top ten. The two kinds of corpus converge from about the 1,000th character.
-
In Chinese Wikipedia, 35.8%. In Jun Da's corpus of fiction and non-fiction, 41.8% — text with more dialogue leans harder on its top characters. The share depends on the corpus, but over a third holds in both.
-
Roughly yes, but not strictly. Prioritise high-frequency characters, but also consider semantic grouping (kinship terms together: 父, 母, 兄, 弟, 姐) and radical patterns (月-component: 明, 朋, 肝). HSK curricula balance frequency with cognitive efficiency.
-
Depends on use case. Not needed for daily reading. But essential for classical literature, medicine, law, personal/place names. E.g. 鼎 and 鼐 are low-frequency but appear often in historical texts.
-
Simple summation: add up the shares of the top N characters. In Chinese Wikipedia the top three — 的, 年, 为 — are 2.72%, 1.43% and 0.87%, a cumulative 5.02%: knowing those three, you recognise about 5% of the characters in the text.
-
Mostly the same — most characters map 1:1 (简龙 ↔ 繁龍). But edge cases exist: some simplified characters merge multiple traditional ones (干 = 乾/幹/干), which can shift rankings. This tool uses modern simplified corpora.
-
We have not measured regional corpora. What is clear is that genre matters more than region: in the same Wikipedia data 我 ranks 465th, while in dialogue-heavy corpora it is usually in the top ten. Counted before conversion, traditional forms such as 為 and 後 rank separately.
-
It affects IME design, not actual character frequency. Pinyin IMEs surface high-frequency characters first; Wubi IMEs encode by radical — but the actual characters people type remain high-frequency (because daily expression needs drive frequency).
-
The current figures come from the 1 September 2026 dump of Chinese Wikipedia. The top of the ranking is stable; when we re-count, the new dump date will be shown here.
Pick up where you left off
Stored only in this browser — never sent to our servers.