Readability checker
This readability checker scores any passage you paste on four measures: Flesch Reading Ease, Flesch-Kincaid grade, Gunning Fog and SMOG. It then lists your longest sentences, which is usually where a reader slows down. It's built for chapters and manuscripts, it runs in your browser, and nothing you paste is uploaded. A plain 54-word paragraph of writing advice scores 91 for ease and grade 3.9 on Flesch-Kincaid. The first 246 words of Pride and Prejudice score 71 and grade 9.4. Results appear once you've pasted 30 words, but a few paragraphs give steadier numbers.
Paste at least thirty words to see the scores.
How to use the readability checker
The tool starts at 30 words, but short samples swing wildly.
It's the headline, and the meter under it shows where you sit.
They should land within a few grades of each other. A wide gap points to unusual vocabulary or very long sentences.
Stop when the writing sounds right, not when a number is hit.
Take out headings, page numbers and bracketed notes before you paste, because they count as words and sentences. If you want a clean word count first, the word and page count tool will give you one.
How the readability checker calculates each score
Words are runs of letters, numbers, apostrophes and hyphens. A sentence ends at a period, exclamation mark, question mark or ellipsis followed by a capital letter or opening quote, with abbreviations such as Mr, Dr, etc and e.g. kept intact. Syllables come from a rule of thumb that counts vowel groups and adjusts for silent endings, plus a short list of exceptions. It's approximate, and it miscounts in both directions. Words ending in a consonant plus "le" (table, little, simple) come out as three syllables instead of two, and words like wanted, started and waited come out as one instead of two. That's one reason two tools disagree on the same text.
206.835 minus 1.015 times words per sentence, minus 84.6 times syllables per word. Higher is easier. Rudolf Flesch published it in 1948. The tool rounds it to a whole number and shows it from 0 to 100; when a text scores outside that range, a note under the composition panel gives the raw figure.
0.39 times words per sentence, plus 11.8 times syllables per word, minus 15.59. The result reads as a US school grade. The formula traces to the 1975 US Navy report that derived new readability formulas by Kincaid and colleagues. The tool shows one decimal and never goes below 0.
0.4 times the sum of words per sentence and the percentage of complex words. Robert Gunning devised it in 1952. A complex word here has three or more syllables, with two exclusions: capitalized words that don't open a sentence (names), and words that reach three syllables only through an -ing, -ed or -es ending. The name check is approximate, so a name at the start of a sentence still counts.
1.0430 times the square root of (words of three or more syllables x 30 / sentences), plus 3.1291. G. Harry McLaughlin proposed it in 1969, and it was designed for 30 sentences. The tool scales your count to 30 so shorter samples still produce a number, but the result is rough. Unlike Fog, SMOG counts every word of three or more syllables, names included.
How to read the numbers
The meter bar changes style below 50 and again below 30. The line under the score uses these bands:
| Flesch score | The tool says |
|---|---|
| 90 to 100 | Very easy, around 5th grade |
| 80 to 89 | Easy, conversational, around 6th grade |
| 70 to 79 | Fairly easy, around 7th grade |
| 60 to 69 | Plain English, around 8th to 9th grade |
| 50 to 59 | Fairly hard, 10th to 12th grade |
| 30 to 49 | Hard, university level |
| Under 30 | Very hard, postgraduate |
The "Suits" line maps the Flesch-Kincaid grade, and it's this tool's own rough mapping: up to 3, early readers (ages 6 to 8); over 3 to 5, chapter books (8 to 10); over 5 to 7, middle grade (10 to 13); over 7 to 9, young adult and most trade fiction; over 9 to 12, general adult nonfiction; above 12, academic and professional readers.
Three plain rules. A score measures sentence length and word length, not quality. Aim for the reader you have in mind, not for one number. And compare chapters of your own book against each other, because that comparison is far more useful than any absolute figure.
Worked examples with real numbers
Here's the sample: "Most first drafts carry more words than the story needs. When you read a chapter aloud, you will hear where a sentence stalls, and that is the place to cut. Keep the verbs that do the work. Trim the rest, and the reader will move through the page without noticing the effort behind it."
It has 54 words, 4 sentences and 65 syllables. That's 13.5 words per sentence and 1.204 syllables per word.
- Flesch: 206.835 minus 13.70 minus 101.83 = 91.3, shown as 91.
- Flesch-Kincaid: 5.27 plus 14.20 minus 15.59 = 3.9.
- Fog: the only three-syllable word is "noticing," which is excluded as an -ing word, so complex words are 0. 0.4 x 13.5 = 5.4.
- SMOG: one word of three syllables, so 1.0430 x the square root of (1 x 30 / 4) plus 3.1291 = 6.0.
The tool shows 91, 3.9, 5.4 and 6.0, the same as the arithmetic.
A single 56-word sentence from an imagined committee report scores Flesch 0 (the raw value is minus 7.1), grade 28.2, Fog 32.4 and SMOG 25.3. Split into four sentences totaling 34 words, it scores Flesch 59, grade 7.2, Fog 11.6 and SMOG 10.7. Sentence length moved the score further than any word choice, and the cut also removed 22 words.
Each sample is the first paragraphs of a Project Gutenberg book, up to about 200 words.
| Book (opening) | Words | Sentences | Flesch | Flesch-Kincaid | Fog | SMOG |
|---|---|---|---|---|---|---|
| The Tale of Peter Rabbit | 224 | 11 | 78 | 7.5 | 9.4 | 8.3 |
| Alice's Adventures in Wonderland | 253 | 6 | 50 | 16.8 | 19.2 | 12.5 |
| Pride and Prejudice | 246 | 10 | 71 | 9.4 | 12.4 | 11.6 |
| Moby Dick | 201 | 8 | 52 | 12.3 | 14.6 | 13.2 |
Peter Rabbit, a picture book, lands on "young adult and most trade fiction," because the formulas count commas and long read-aloud sentences, not how a parent reads them. Alice has just 6 sentences in 253 words, and one 106-word sentence drives the grade to 16.8, which shows how a small sample misleads. Pride and Prejudice averages 24.6 words a sentence and sits near grade 9. None of this says one book is better than another.
A 43-word run of clipped dialogue ("Go." "No." "Now.") scores Flesch 100 (raw 110.3) and grade 0.0, at about two words a sentence. That tells you nothing about how easy the scene is to read. For dialogue-heavy fiction, score the narration paragraphs separately.
Where a readability checker misleads you
- Scoring under 100 words, or fewer than 5 to 10 sentences. SMOG in particular is a guess on a short sample.
- Trusting one number. Flesch ease and the grade levels all measure sentence length and word length, so their agreement isn't independent proof.
- Reading "grade 7" as a promise that a seventh grader will understand the book. The formulas know nothing of meaning, structure or interest.
- Chasing the easiest possible score. Short sentences and one-syllable words, all the way down, read monotone. Mix lengths.
- Scoring poetry, lists, scripts or recipes. The sentence detector needs full stops.
- Pasting with headings or notes still in. They become words and sentences.
- Comparing scores across tools. Each counts syllables and sentences its own way, and the same text can land several points apart.
- Taking the passive percentage as exact. It's the share of sentences containing a form of "be" plus a past participle, a loose rule that also catches adjectives such as "broken." The tool flags anything above a quarter of sentences, which is its own threshold. A free line editing tool lists the actual passives.
What readability targets suit which kind of book
These are an editor's rules of thumb built from the tool's bands and the table above. They aren't industry standards.
Fiction
Our two most readable samples, Peter Rabbit and Pride and Prejudice, land at Flesch 78 and 71, while Moby Dick (52) and Alice (50) score lower and can still be right for their voices. Score narration without dialogue. Then look at your longest sentences and whether a run of them sits inside an action scene.
Memoir and narrative nonfiction
Conversational is the point, so aim for the band a reader expects from a person talking. Check chapters against each other, because memoir tends to drift into long reflective sentences. A structural problem comes before a sentence problem; if the shape is off, start with a developmental edit.
Business and nonfiction
Readers skim, so plain English at 60 to 69 on the Flesch scale is the tool's own band for this work. Technical terms raise the grade. Define them and break up the sentences around them rather than removing the terms. Pair the score with a line edit.
Children's and early readers
The "Suits" line is a rough age band, not a book level. Picture books are read aloud, so sentence flow matters more than the score, as Peter Rabbit shows. For chapter books, grade and sentence length are useful checks, but test with real readers too.
How to raise the score without flattening your voice
Split the longest sentence
Start from the "Longest sentences" list. One split often moves the score more than twenty word swaps. The federal guidance on writing short sentences makes the case.
Swap the long word for the plain one
Only when the plain word is just as exact. Keep the technical term your reader needs.
Cut the passive where it hides the actor
Some passives are right. Treat the percentage as a pointer to where to look.
Vary the rhythm
Short, medium, long. A uniform score of 80 can read flat.
Read it aloud
The ear catches what a number can't. Then paste again.
Related guides on the Author Resources hub
- How to Self-Edit Your Book: the order of passes and what each one catches.
- The Types of Book Editing, Explained: developmental, line, copy and proofreading side by side.
- How to Write a Book: drafting habits that make the editing easier.
When a score points to a line edit
If the longest sentences and the grade levels say the prose needs work and you'd rather have a person do it, our line edit goes sentence by sentence for rhythm, clarity and voice, as tracked changes you accept or reject. Our published rate is $0.027 a word, so a 50,000-word memoir comes to $1,350 (0.027 x 50,000), with a $450 minimum per pass. We read a sample before quoting and will say if the manuscript needs an earlier stage. Writing for younger readers? See our children's book writing services. The full list is on our published editing rates page.
Our published rate is $0.027 a word, so a 50,000-word memoir comes to $1,350 (0.027 x 50,000), with a $450 minimum per pass.
Questions we get asked
What is a readability checker, and what does it measure?
It scores how hard a passage is to read from sentence length and word length. This readability checker gives you Flesch ease, three grade levels and your longest sentences. It doesn't measure meaning, quality or interest, so a high score never means the writing is good.
What is a good readability score?
It depends on the reader. On the Flesch scale, 60 to 69 is plain English, 80 and above is easy, and below 50 is hard going. Match the score to the person you're writing for rather than to a target. A technical manual and a beach novel shouldn't land in the same place.
What reading level should my book be?
There's no official level for a trade book. Our four classic samples run from grade 7.5 to 16.8, and adults read all of them. Aim at the reader you have in mind, use the targets section and the classics table above as rules of thumb, and test with real readers if the audience is young.
How do I check readability in Word?
Word shows Flesch Reading Ease and Flesch-Kincaid Grade Level in its Readability Statistics window once you turn on that option and run the spelling and grammar check. Microsoft's guide to readability statistics in Word has the steps for your version. Our tool adds Fog, SMOG and a list of your longest sentences.
Is this readability checker free and private?
Free, with no sign-up. The page's script scores your text in the browser and doesn't upload it, which matters when you're pasting an unpublished chapter. We don't store what you paste.
Which readability formula should I trust?
None alone. Flesch ease and the three grades all lean on sentence and word length, so use them as one signal. Look at the spread between the grades, read the longest sentences and then read the passage. SMOG is common in health writing, and it's the roughest of the four on a short sample.
Why is my score different on different tools?
Tools count syllables and sentences differently. Ours uses a rule of thumb for syllables and treats a period, exclamation mark, question mark or ellipsis followed by a capital as a sentence end. A different splitter or syllable dictionary can move the score by a few points, so compare scores from one tool only.
Does the readability checker work for dialogue, poetry or children's books?
Poorly on dialogue and poetry. Short clipped lines produce an artificially easy score; our dialogue sample scored 100. For children's books, the "Suits" line is a rough guide, so read aloud and test with real readers. Score narration and prose chapters apart from dialogue.
Keep going with the next tools
Once you've tightened the sentences, try the free line editing tool for adverbs, filter words and repeated phrases, the word and page count tool for length, and the style sheet builder to record your spelling and punctuation choices.
Tools will only take you so far
Send what you have. Even a half-written manuscript is enough for a reader to tell you what it needs.