Skip to content
TextLab

Free average word length calculator

Average word length in letters and syllables

TextLab's average word length calculator reports the mean and the median letters per word for anything you paste, alongside syllables per word, words per sentence and the share of words running to seven letters or more. Free, with no account to make, and it remeasures as you type. A histogram shows every word length from one letter to fifteen, and the ten longest words are listed by name — which is usually more useful than the average that summarises them.

  • 100% free
  • No signup
  • Mean and median
  • Length histogram
  • Longest words listed
0.00letters per word0 letters across 0 words

0
Median letters
half your words are shorter than this
0.00
Syllables per word
sound, not spelling
0.0
Words per sentence
the other axis of difficulty
0
Letters per sentence
how much there is to hold in mind

Where the word lengths fall

A bar per word length will appear here. English is lopsided: the tall bars sit at three and four letters, and the mean is dragged rightwards by a long tail of rare, long words.

How to measure average word length

By the third step the number has names attached to it.

  1. Paste the text you want characterised

    This works on anything from a single abstract upwards, and it is most revealing on two texts compared against each other — the same argument written for colleagues and for customers, or a translation set beside its original. The mean settles quickly, so a few hundred words is plenty.

  2. Compare the mean against the median

    The headline mean is pulled to the right by a small number of very long words, while the median sits where half your words are shorter and half longer. In ordinary English the median is typically a letter or more below the mean, and the size of that gap tells you whether your text is uniformly plain or plain with occasional five-syllable interruptions.

  3. Look at the tail of the histogram

    The bars for one to fifteen letters show the actual distribution rather than a summary of it, and the informative part is the right-hand end. A visible tail past nine letters is what makes prose feel dense, and the list of your longest words beside it names the specific offenders with their syllable counts, which is usually more actionable than any average.

Technical specifications

What counts as a letterUnicode letters only; digits, punctuation, hyphens and apostrophes are excluded, so state-of-the-art is 14 letters and don't is 4
Mean and medianBoth reported, because the distribution is skewed — the median is typically a letter or more below the mean in English prose
HistogramBars for word lengths 1 through 14 with everything of 15 letters or more pooled into the final bar, each with its count and share
Long-word threshold7 letters or more, the cut-off long-word ratios and the Scandinavian LIX index use
Longest wordsThe ten longest distinct words in the text, each with its letter count and its estimated syllables
Typing-test conversionThe same text measured the way a typing test measures it, at a flat 5 characters to the word regardless of spaces
Register bandsFour reference bands from spoken English below 4.2 letters per word to technical and legal writing above 5.4
PriceFree, with no account, no watermark and no cap on how many drafts or translations you compare

Frequently asked questions

What is the average word length in English?

Around 4.7 to 5.1 letters per word for written English, depending on whose corpus you count. Conversation and dialogue run shorter, nearer 4.2, because spoken English leans on a small core of short, frequent words. Academic and legal prose run longer, often past 5.4. If your text lands outside 4 to 6 letters per word, something specific is going on: a glossary, a list of names, a heavily technical register, or a text that is not really English prose.

Why is my median lower than my mean?

Because word lengths are skewed, not symmetrical. There is a floor at one letter and no ceiling, so the handful of long words in any text pull the average up while leaving the middle of the distribution where it was. This is normal and universal — in most English prose the median is three or four letters and the mean is closer to five. A median that matches the mean would be the surprising result, and usually means an unusually uniform vocabulary.

Do numbers, hyphens and apostrophes count towards word length?

Letters only. In state-of-the-art the hyphens are ignored and the fourteen letters are counted; in don't the apostrophe is skipped and four letters are counted; a bare figure such as 1,500 counts as a word with zero letters and is therefore left out of the length figures entirely, though it still appears in the word count. That keeps the average measuring what it claims to measure — the length of words — rather than the length of tokens.

Is a longer average word length better or worse writing?

Neither, and treating it as a quality score is the main way this measure gets misused. A pharmacology paper needs its long words; a children's book needs short ones; a press release with a 5.8-letter average is probably hiding behind abstraction. What the figure does tell you reliably is register — how formal and how specialised the text reads — and that is genuinely useful when you are checking whether a piece matches the audience it was written for.

How does word length feed into readability formulas?

Two of the standard formulas use it directly. Coleman-Liau and the Automated Readability Index both count letters per word instead of syllables per word, which was a deliberate choice: in 1967 and 1975 a machine could count characters trivially and syllables not at all. The trade-off is that letters cannot distinguish a familiar long word from an unfamiliar one, so both formulas rate technical jargon that happens to be short as easy. The syllable-based formulas make the opposite mistake.

What is the average number of syllables per word?

About 1.4 to 1.5 in general English prose, and 1.7 to 1.8 in academic writing. It is a narrow range compared with sentence length, which is precisely why readability formulas that use it apply a large multiplier — a shift from 1.4 to 1.6 syllables per word is a substantial change in register even though the number barely moves. Syllables and letters usually agree about a text, but not always: monosyllabic words can be long, as strengths and thoughts demonstrate.

Why does my translation have a different average word length?

Because languages build words differently. German and Finnish compound freely, so a concept English writes as three short words becomes one long one and the average jumps; the Romance languages tend to longer words than English for the same content; Chinese written in characters is not comparable at all. A translation whose average word length is a letter above the original is usually accurate rather than wordy — compare sentences and syllables as well before concluding the translator inflated anything.

Why English words are short, and what it means when yours are not

English has an unusually short average word for a European language, and the reason is historical. Its most-used vocabulary — the words for eating, sleeping, working, hitting and dying — is Germanic and mostly one syllable, while the Latin and French layer that arrived after 1066 supplies the abstract, formal and technical register on top. That gives English something most languages lack: a genuine choice between a short word and a long one for very nearly the same idea. Buy or purchase, end or terminate, ask or enquire, help or assistance. Nothing forces the long version, which is exactly why an elevated average word length reads as a decision rather than a constraint. There is also a statistical pressure keeping the common words short, first described by George Zipf in the 1930s: the more often a word is used, the shorter it tends to be, in every language anyone has checked.

The distribution matters more than the mean, which is why the histogram above is the main event. Word lengths pile up at three and four letters and then trail away to the right with no upper bound, so the mean sits above the median and is dragged around by the rarest words in the text. Two documents can share an average of 4.9 letters and read completely differently: one uniformly mid-length, the other short words punctuated by twelve-letter abstractions. It is the tail that makes prose feel heavy, and the tail is visible in the bars rather than in the average.

The measure has a practical afterlife in typing and in readability. Typing speed is reported in words per minute, but the word of a typing test is a fixed five characters — an arbitrary convention adopted so that a test on a paragraph of short words is not easier than one on a paragraph of long ones, and the reason the panel above shows what a typing test would call your text. In readability, two of the standard formulas count letters rather than syllables for the same reason: letters were countable by machine decades before syllables were. To see those formulas run over your text, the readability score panel computes six of them at once; the syllables-per-word figure above is worked out by the routine driving the syllable counter, and the Flesch Reading Ease calculator shows exactly how much of a score your word length is responsible for. If you want the raw character totals instead of averages, the character counter has them with and without spaces.

Your text is measured where you pasted it

Letter counting, syllable estimation and the histogram are all done by the page itself, so nothing you paste is uploaded or recorded — which is worth knowing if you are comparing a confidential translation against its source, or checking a colleague's draft without telling them. There is no account, no history and no saved document; closing the tab removes every trace of the text from the machine.