Free tool Readability checker with long sentences marked
Paste a draft to get six standard English readability scores and see exactly which sentences run long. Chinese has no widely accepted formula, so Chinese text gets plain indicators instead: sentence length, long-sentence share, paragraph length and how many characters fall outside the common-character list.
- Six English formulas
- Chinese indicators too
- Runs locally
Check readability
Paste body text; headings and bullet lists without punctuation count as separate sentences. Each sentence is checked as English or Chinese on its own, so mixed text works.
An article, help page, email or product description. At least 100 words gives the formulas something to work with.
Sentences longer than this are highlighted. 25 words is the limit in the GOV.UK style guide.
In characters. There is no standard for Chinese; pick what suits your readers.
Flesch Reading Ease
77.6
0–100; higher is easier.
Flesch-Kincaid Grade
5.4
US school grade.
Gunning Fog
8.5
Years of schooling.
SMOG
9.7
US school grade.
Coleman-Liau
8.4
US school grade; letters, not syllables.
ARI
5.8
US school grade; characters per word.
Words per sentence
11.8
English sentences only.
Syllables per word
1.39
Estimated, see the rules below.
Complex words
9.4%
Three or more syllables (Gunning's rule).
Long sentences
11.1%
Share of English sentences over the limit.
Where the Flesch score falls
Bands from Rudolf Flesch, The Art of Readable Writing (1949).
90–100
Very easy
5th grade
80–90
Easy
6th grade
70–80
Fairly easy
7th grade
60–70
Standard
8th–9th grade
50–60
Fairly difficult
10th–12th grade
30–50
Difficult
College
0–30
Very confusing
College graduate
Characters per sentence
—
Chinese sentences only.
Long sentences (Chinese)
—
Share of Chinese sentences over the limit.
Longest stretch without punctuation
—
In characters; the longest run a reader gets no pause.
Characters per paragraph
—
Average length of a paragraph (non-empty line).
In the 2,500 common characters
—
Share of all Chinese characters, counting repeats.
In the full 3,500 list
—
2,500 common + 1,000 less common characters.
Notes
- CheckSentences over the limit (25 words for English, 40 characters for Chinese): 1. Each one is highlighted below.
- NoteSMOG was designed for a 30-sentence sample; with 9 sentences the score is extrapolated.
Sentence by sentence
Hover a sentence to see its length.
Long sentences
| # | Sentence | Length |
|---|---|---|
| 4 | Readers scan a page before they decide to read it, and a sentence that carries three ideas, two conditions and a list of exceptions forces t… | 40 words |
Version v1.0Updated Update history
What the scores mean
Every formula here measures surface features: sentence length and word length. None of them knows whether the text makes sense. Use them to find heavy passages, then judge those passages yourself.
Flesch Reading Ease
A score from roughly 0 to 100. Higher is easier. Flesch put 60–70 at "standard", the level of plain consumer writing; academic papers often score below 30. The band table under the scores shows where your text falls.
Grade levels
Flesch-Kincaid, Gunning Fog, SMOG, Coleman-Liau and ARI estimate the US school grade needed to follow the text. They weigh different things (syllables, long words, letters, characters), so a spread of two or three grades between them is normal.
Long sentences
The most direct thing to fix. GOV.UK's style guide asks writers to break up sentences over 25 words. The marked text and the table show each one with its length.
Chinese indicators
Characters per sentence, the share of long sentences, the longest stretch with no punctuation, paragraph length and how many characters are outside the 3,500 common characters. Each one is a count you can check, not a grade.
Readability scores do not measure how long people actually spend on a page or whether they finish it. Compare drafts with time on page and scroll depth once the page is live.
Make a draft easier to read
The quickest gains come from the marked sentences, not from chasing a score.
-
1
Paste the draft
Body text only. Each line without end punctuation, such as a heading or list item, counts as its own sentence.
-
2
Set your limit
Keep 25 words for general web copy, or go lower for help articles and audiences reading in a second language.
-
3
Fix the marked sentences
Split the orange and red ones, or move a list out of the sentence. The scores update as you edit.
-
4
Check the scores last
Compare before and after. If the grade level is far above your audience, look for long words that have shorter, common alternatives.
Six formulas, as published
W = words, S = sentences, Y = syllables. Coefficients follow the original papers; where the popular Python library textstat differs, it is noted below.
- Syllables are estimated with spelling rules plus a short exception list. On 116 words the rules had never seen, 94% matched the CMU Pronouncing Dictionary; textstat looks words up in that dictionary, so expect small differences on rare words
- Complex words for Gunning Fog: three or more syllables, not counting capitalised words mid-sentence (proper nouns), hyphenated compounds, or words that only reach three syllables through -es or -ed. SMOG counts every word of three or more syllables
- Coleman-Liau uses 0.0588 as in the 1975 paper (textstat uses 0.058). ARI counts letters and digits as characters, as Smith and Senter did (textstat also counts punctuation)
- Numbers count as one-syllable words. A sentence ends at . ! ? and line breaks, but not after abbreviations such as Mr. or e.g.
Flesch Reading Ease
206.835 − 1.015 × (W ÷ S) − 84.6 × (Y ÷ W)
Flesch, 1948
Flesch-Kincaid Grade
0.39 × (W ÷ S) + 11.8 × (Y ÷ W) − 15.59
Kincaid et al., 1975
Gunning Fog
0.4 × [(W ÷ S) + 100 × (complex words ÷ W)]
Gunning, 1952
SMOG
1.0430 × √(polysyllables × 30 ÷ S) + 3.1291
McLaughlin, 1969
Coleman-Liau
0.0588 × L − 0.296 × S₁₀₀ − 15.8
L, S₁₀₀ = letters and sentences per 100 words; Coleman & Liau, 1975
ARI
4.71 × (characters ÷ W) + 0.5 × (W ÷ S) − 21.43
Smith & Senter, 1967
The example text, scored
The default text is 106 words in 9 sentences, with 147 syllables, 12 words of three or more syllables and 481 letters. One sentence runs to 40 words.
| Formula | This tool | textstat | Why they differ |
|---|---|---|---|
| Flesch Reading Ease | 77.6 | 77.6 | Same counts, same formula |
| Flesch-Kincaid Grade | 5.4 | 5.4 | Same |
| Gunning Fog | 8.5 | 8.5 | Same here; complex-word rules differ in general |
| SMOG | 9.7 | 9.7 | Same |
| Coleman-Liau | 8.4 | 8.0 | Coefficient 0.0588 vs 0.058 |
| ARI | 5.8 | 6.7 | textstat counts 20 punctuation marks as characters |
Flesch 77.6 falls in the "fairly easy" band. The 40-word sentence in the second paragraph is marked red because it is more than 1.5 times the 25-word limit; splitting it into three sentences ("…decide to read it. A sentence that… forces them to stop. They go back to the start…") clears the only warning and lifts the Flesch score to 79.7.
Where readability scores mislead
The formulas are old, simple and useful. They break when asked to do more than they were built for.
Writing for the score
Chopping every sentence in half and swapping in one-syllable words raises the number but can make text choppy and vague. Fix the sentences that are hard, not all of them.
Scoring a few lines
Formulas average over sentences. A 30-word paragraph can swing two grades with one edit. Check full sections of 100 words or more.
Pasting lists and headings
Lines without end punctuation count as short sentences and make text look easier than it is. Score the prose, or accept that lists lower the averages.
Applying English formulas to Chinese
Syllable and letter counts mean nothing for Chinese. This tool scores Chinese sentences separately with character-based indicators instead.
From easy to read to actually read
A readability score is a prediction. TapCub shows what readers do: how long they stay, how far they scroll and where they stop.
Page reports
Visitors, views, average time, bounce and exit rate for every page, with high exit rates highlighted.
Learn moreHeatmaps and scroll depth
Click and scroll maps per page show how far down a long article people actually get.
Learn moreAI answers from your help articles
TapCub live chat answers from the knowledge base you publish and names the article it used. Clear articles make clearer answers.
Learn moreReadability checker questions
Still have a question?
Chat with the team behind TapCub — we usually reply within a few hours on business days.
What is a good Flesch Reading Ease score?
It depends on who reads the text. Flesch called 60–70 "standard" and 70–80 "fairly easy"; most general web copy aims somewhere in that range, while technical documentation often lands lower. Compare a draft with your own best-performing pages rather than with one fixed number.
Why do the grade levels disagree?
Each formula uses different inputs. Flesch-Kincaid and SMOG count syllables, Coleman-Liau and ARI count letters, and Gunning Fog counts complex words. Texts with many short but uncommon words, or long but familiar ones, separate them. Look at the range, not one number.
Why is my score different in another tool?
Tools split sentences and count syllables in different ways, and some use slightly different coefficients. This page shows its own counts, and its test suite compares them with the Python library textstat and the original papers.
How are syllables counted?
By spelling rules (vowel groups, silent e, -ed and -es endings, -le, and common vowel pairs that split, as in "period") plus a short exception list. It is an estimate: on words outside the exception list it matched the CMU Pronouncing Dictionary 94% of the time.
Can it check Chinese text?
Yes. Chinese sentences are not run through the English formulas, because syllables and letters do not apply. They get characters per sentence, the share of long sentences, the longest stretch without punctuation, paragraph length and coverage of the 3,500 characters in the 1988 Modern Chinese Common Characters List.
What counts as a long sentence?
More than 25 words by default, the limit GOV.UK sets for its writers. You can change it. Sentences more than 1.5 times the limit are marked red.
How is this different from the word counter?
The word counter tells you how long a text is. This tool tells you how hard it is to read and which sentences make it so.
Is my text uploaded?
No. All scoring happens in your browser. The text is not saved, not put in the page address and not sent to TapCub.
Update history
New features, fixes and data refreshes for this tool, newest first.
-
· v1.0
LaunchedFirst release.
Missing something or found an error? Tell us — requests that come up often go into the next update. All tool updates
See whether people finish your pages
Add TapCub to your site to see time on page, scroll depth and exits for every article you edit.