Guide
Ready to try your Chinese keyboard?
Start with a short HSK 1 typing session. No signup is required.
Practice HSK 1 vocabularyFor Chinese typing, characters per minute (CPM) is usually easier to interpret than words per minute. A result near 27 CPM is the median among returning TypeChinese typists in our May–August 2026 benchmark; 42 CPM is around the 75th percentile, and 57 CPM is around the 90th percentile.
Those numbers are useful reference points, not universal standards. Your input method, text difficulty, accuracy, and familiarity with the passage all affect the result.
| Percentile | Speed |
|---|---|
| 25th | 15 CPM |
| Median | 27 CPM |
| 75th | 42 CPM |
| 90th | 57 CPM |
A practical interpretation is:
The sample covers 66,771 valid completed practices from 3,437 returning typists between May 1 and August 2, 2026. A returning typist needed at least three valid completed practices.
We excluded pasted input, implausible speeds outside 1-300 CPM, sessions shorter than five seconds, and sessions longer than one hour. To stop frequent users from dominating the result, we calculated one median speed per typist before finding the percentiles.
This is a self-selected sample of people using TypeChinese, not a representative survey of every Chinese typist. It is most useful for comparing your progress with other people who intentionally practice Chinese typing.
English typing tests commonly use WPM, with a “word” standardized as five characters. Chinese does not use spaces between every word, so software may count Chinese words differently.
CPM avoids that ambiguity: 150 counted characters in five minutes is 30 CPM. When comparing results from different sites, first check whether their “WPM” actually means characters per minute.
For a completed passage, CPM = passage characters ÷ elapsed minutes, rounded to the nearest whole number. The passage count includes its punctuation; it is not the number of Pinyin letters, Zhuyin keys, or Cangjie roots pressed. Time spent finding candidates, correcting input, or reading a hint reduces the completed result.
For example, completing 90 passage characters in 180 seconds gives 90 ÷ 3 = 30 CPM. Completing the same 90 characters in 150 seconds gives 90 ÷ 2.5 = 36 CPM.
TypeChinese reports accuracy = passage characters ÷ (passage characters + recorded mistakes) × 100, rounded to a whole percent. With 90 characters and five recorded mistakes, 90 ÷ 95 × 100 rounds to 95%. Correcting mistakes does not erase the fact that they occurred. This is the practice score’s formula, not a count of every physical keystroke or every edit inside the IME’s candidate window.
Other tests may count punctuation differently or subtract errors from speed. Compare their scoring rules before converting or comparing a result.
Pinyin and Zhuyin depend on pronunciation and candidate selection. Cangjie, Quick, and Wubi depend more on recognizing character shapes and recalling codes. Do not treat a score from one method as a perfect prediction for another.
Common HSK vocabulary is easier than names, formal news, or unfamiliar literary text. Compare like with like whenever you track improvement.
A high raw speed with many corrections is not fluent typing. Track accuracy alongside CPM and aim for at least 95% before deliberately pushing speed.
Repeating a passage is useful practice, but the second attempt will usually be faster. Label repeated tests so you do not confuse familiarity with general improvement.
Choose one practice category, script, and input method. Complete three short sessions per week, then compare your weekly median rather than your single fastest result.
| Illustrative week | CPM from three new passages | Median CPM |
|---|---|---|
| Week A | 24, 36, 30 | 30 |
| Week B | 28, 31, 32 | 31 |
Week A’s best score is 36, but week B’s typical score is slightly higher. That is why a fastest-run comparison can tell a different story from a median. These figures demonstrate arithmetic; they are not measured learning outcomes.
Practice a difficult passage repeatedly when it helps, but label those runs as rehearsal. Compare fresh news with fresh news, vocabulary with vocabulary, and similar hint use with similar hint use. The historical benchmark pools TypeChinese practice conditions; it does not supply a separate norm for each IME, device, or category.
For practice between tests, follow the seven-day Pinyin typing plan. To track copying and proofreading separately in a text editor, use the recording sheet in Measuring Chinese Typing Improvement; its timing and scoring rules differ from TypeChinese's practice scores.
The percentiles describe typists’ median speeds, not the fastest sessions and not a representative population survey. A single 57 CPM run does not establish that your usual speed is in the top 10% of all Chinese typists. The data also does not establish a hiring requirement, an examination pass mark, or how much a beginner will improve in two weeks.
Use the figures as context for a series of comparable results. This article retains the original May–August 2026 snapshot; the October update clarifies measurement and does not claim a new data sample.
Yes. In this sample, 30 CPM is slightly above the median returning typist. Accuracy and consistency still matter more than crossing an exact number.
About 42 CPM reaches the 75th percentile in this sample. Around 57 CPM reaches the 90th percentile.
There is no fixed schedule. In the first few weeks, fewer pauses and better accuracy are stronger signs of progress than raw speed. Measure under consistent conditions and look for a trend across several sessions.