CPM vs WPM: Reading Speed in Chinese, Japanese, and English
WPM counts words; CPM counts characters. Learn why the two units don't convert and how to benchmark your reading speed within each language.
WPM counts words; CPM counts characters. They measure different scripts, so no fixed conversion exists between them — a 250 WPM result in English tells you nothing about your Chinese or Japanese reading until you measure those languages in CPM.
This matters more as reading tools go multilingual. If you read in more than one language, or you are learning to, the numbers on the screen can quietly mislead you: a Chinese result expressed in "words per minute," an English score compared against a Japanese one, a conversion someone dividing CPM by two invented on a forum. Here is what each unit actually counts, why the gap between them is structural, and the only two comparisons that stay honest.
What WPM and CPM each count
WPM — words per minute — fits languages that put spaces between words. English word counts are well defined: whitespace separates units, and decades of reading research report pace in words — the unit unpacked in what is WPM in reading — which is why the widely cited adult silent-reading average sits around 238–260 WPM (Brysbaert et al., 2019).
CPM — characters per minute — fits scripts written without spaces. Chinese text runs character after character, and Japanese mixes kanji and kana the same way. The reader's eyes work through characters, so pace is counted in characters: 字/分钟 in Chinese, 文字数/分 in Japanese.
The counting rules matter as much as the unit. In LumenRead, Chinese-primary text counts Han characters and Japanese-primary text counts Han characters plus kana — punctuation and whitespace are excluded in both cases, just as word counts exclude them. A CPM number quoted without its counting rule is not comparable to anything, not even to another CPM number.
Why the two units don't convert
The units look interchangeable — both are "things per minute" — but the things are different sizes of language:
| Unit | What one unit is | Typical load per unit |
|---|---|---|
| WPM (English) | One whitespace-delimited word — average about 5 letters plus a space | a word, sometimes two function words |
| CPM (Chinese) | One Han character — roughly a syllable and often a morpheme | less than an English word on average |
| CPM (Japanese) | One kanji or kana character — kana carry grammar, kanji carry meaning | varies widely with the kanji–kana mix |
One English word usually takes about one and a half to two Chinese characters to say in translation. That single fact breaks every conversion: a character is a smaller package than a word, so a CPM number is naturally larger than the WPM number for the same text — and by a ratio that changes with every book, because translation length is not fixed.
The same sentence, counted in two units: fewer large packages or more small ones. Illustration, not measured data.
The one bridge that works: same book, same time
There is a comparison that stays honest, and it is not a formula — it is a book. Translation publishers know the effect well: the Chinese edition of an English novel typically runs about one and a half to two characters for every English word, and the Japanese edition varies with its kana–kanji mix. Cross-language reading research points the same direction: when people read standardized translated texts, total reading times for equivalent content come out broadly similar across languages (Trauzettel-Klosinski & Dietz, 2012).
Put those together and you get a workable estimate, not a conversion:
- If an English edition has 90,000 words and the Chinese edition about 160,000 characters, and you read both in roughly the same wall-clock time, your Chinese CPM on that book will land near 1.8 times your English WPM.
- The multiplier — 1.5, 1.8, 2.1 — belongs to that book pair, not to you. Pick another title and it moves.
That is why reading tools should never "convert" your score between units. The multiplier is a property of the text pair, estimated after the fact, and useless for a test result produced in one session.
What people get wrong
Three mistakes account for most of the confusion:
- Relabeling. Taking an English WPM figure and calling it CPM. The number keeps its size and loses its meaning — it was counting words that do not exist in the Chinese text.
- Dividing or multiplying by a fixed factor. Any fixed factor assumes every text translates at the same ratio. They do not, and the error compounds when the factor is then compared against benchmarks built in the other unit.
- Comparing across languages. "I read 500 in Japanese and 250 in English, so I read Japanese twice as fast." The units differ; only the session inside one language, on comparable material, with a comprehension check, is a fair comparison.
If you want the English benchmarks themselves, they live in average reading speed, average reading speed by age, and the WPM reading speed chart — all of them English charts, labeled as such.
How to benchmark yourself in each language
Treat each language as its own scoreboard:
- Measure one baseline per language on comparable material — reading WPM test for English, the Chinese reading speed test for 中文, the Japanese reading speed test for 日本語. Each reports pace in its own unit, alongside completion and a comprehension check.
- Compare only within a language, over time. Your Chinese CPM this month against your Chinese CPM last month is a real signal. Your Chinese CPM against your English WPM is not a comparison; it is two numbers wearing the same clothes.
- Keep the comprehension check in the loop. Units make speed look comparable across languages; comprehension is what actually makes the sessions comparable.
Bilingual readers usually find their two numbers differ, sometimes a lot — that is a fact about vocabulary exposure, not about intelligence, and it is exactly why the baselines stay separate. How to measure pace without fooling yourself is covered in how to measure reading speed accurately.
Start from your own baseline
The WPM–CPM gap is not a bug to be converted away; it reflects real differences between scripts. Measure each language you read in its own unit, compare against yourself over time, and let the quiz referee every gain. A free reading speed test in any of the three languages takes a few minutes and gives you the baseline the conversions can't.
One scoreboard per language. Illustration, not measured data.
Frequently asked questions
What does CPM stand for in reading?
Characters per minute — the pace unit for scripts written without spaces, primarily Chinese (Han characters) and Japanese (kanji plus kana). WPM, words per minute, is the equivalent for space-delimited languages like English.
Can you convert WPM to CPM?
No fixed conversion exists, because a word and a character are different sizes of language and the ratio changes with every text. For rough estimation only: translated Chinese editions typically run about 1.5–2 characters per English word. Never use that ratio to convert a test score.
What is a good CPM for Chinese reading?
There is no equivalent of English's widely cited 238–260 WPM average — Chinese reading research varies in counting rules and materials, which makes single numbers unreliable. Establish your own baseline on reviewed passages and track it over time within the same counting rule.
Why is a CPM number bigger than a WPM number for the same book?
Because characters are smaller units than words — it takes roughly one and a half to two Chinese characters to carry an English word. More, smaller packages per minute is arithmetic, not faster reading.
How is CPM counted in LumenRead?
Chinese-primary text counts Han characters; Japanese-primary text counts Han characters plus kana. Punctuation and whitespace are excluded, and speed uses active visible reading time only.
Do bilingual readers read at the same speed in both languages?
Usually not. Pace tracks vocabulary exposure and reading history per language, so most bilingual readers hold two genuinely different baselines. Track each language separately and compare yourself to yourself.
Which unit should I track?
The one that matches the script you are reading: WPM for English, CPM for Chinese and Japanese. If you read across languages, keep one scoreboard per language rather than forcing them onto one scale.
Different scripts, separate scoreboards
WPM and CPM are not rivals waiting for a conversion rate; they are honest units for different scripts. Count words where words exist, count characters where they don't, keep the counting rule fixed, and compare yourself within a language over time.
Start with a free reading speed test, set one baseline per language, and let the retests tell the story.