Readability
Score how hard your writing is to read with the two standard Flesch tests. See the Flesch Reading Ease score, its plain-language band, the Flesch–Kincaid grade level, and the average words-per-sentence and syllables-per-word figures behind them, all updating live as you type. The averages tell you whether long sentences or long words are dragging a passage down, so you know which to fix first.
How Readability works
Paste or type the text you want to assess.
Read the reading-ease score and its plain-language band.
Check the grade level for the school year it suits.
Shorten long sentences and simpler words to raise the score.
When to use Readability
- Checking that a landing page or help article reads at a general-audience level before it ships, rather than discovering after launch that it lands at a college grade.
- Aiming a piece at a target school grade — student materials, patient information, public-sector copy — by watching the Flesch–Kincaid figure fall as you simplify.
- Diagnosing why a paragraph feels heavy: the average-words-per-sentence and average-syllables-per-word rows tell you whether long sentences or long words are the culprit.
- Comparing two drafts of the same passage side by side to see which one actually reads more easily rather than which one feels shorter.
- Loosening dense corporate or academic prose until the reading-ease band climbs out of “Difficult” or “Very Confusing” into “Standard”.
- Editing for a second-language audience, where a lower grade level and shorter sentences make the difference between skimmable and unreadable.
- Setting an internal house standard — say, a Flesch Reading Ease of 60 or above for public help content — and checking drafts against it before review.
- Spot-checking a single unwieldy sentence by pasting it alone to see how far it drags the syllable average up.
Examples
Input
The cat sat on the mat and watched the rain.
Output
10 words, 1 sentence, 10 syllables — one per word. Average words per sentence is 10.0 and average syllables per word is 1.00, which puts the Flesch Reading Ease above 100 in the “Very Easy” band and the Flesch–Kincaid grade near zero.
Scores above 100 are real: the Flesch scale is open-ended, and the top band absorbs anything over 100.
Input
The utilisation of superfluous polysyllabic terminology substantially diminishes comprehensibility.
Output
9 words but 35 syllables, so average syllables per word is 3.89. That pushes reading ease deep into negative territory, well inside the “Very Confusing” band, and the grade level past 30 — college-graduate and beyond — even though it is a single short sentence.
Word length, not sentence length, is doing all the damage here; the sentence is shorter than the plain one above.
Input
Local first tools keep your data on your own device. Nothing you type is uploaded to a server. That makes them fast and private.
Output
24 words across 3 sentences, 31 syllables. Average words per sentence is 8.0 and average syllables per word is 1.29, giving a reading ease around 89 in the “Easy” band and a grade level near 3.
Short sentences and mostly one- and two-syllable words are what keep this in the easy range.
Input
A long sentence built from short words versus a short sentence built from long words.
Output
A rambling forty-word sentence of plain words can still score well, while a seven-word sentence of technical terms scores badly — because reading ease weights syllables per word far more heavily than words per sentence.
How Readability works under the hood
Two standard formulas run on every keystroke. Flesch Reading Ease is 206.835 − (1.015 × average words per sentence) − (84.6 × average syllables per word); a higher score is easier. Flesch–Kincaid Grade is (0.39 × average words per sentence) + (11.8 × average syllables per word) − 15.59, and it estimates the US school year needed to follow the text. Both use exactly these coefficients, so the numbers match any reference implementation of the classic tests.
The two inputs are simple ratios. Average words per sentence is the word count divided by the sentence count. Average syllables per word is the total syllable estimate divided by the word count. Notice the 84.6 and 11.8 coefficients: syllables per word swing the result far more than sentence length, which is why swapping a long word for a short one moves the score more than trimming a sentence by a few words.
Words and sentences reuse the shared NeatKit tokenisers. Words are whitespace-separated tokens; sentences close at a full stop, question mark, or exclamation mark followed by whitespace, with a guard list for abbreviations like “Dr.” and “e.g.” and for decimals so “3.5 million” does not split. To keep a fragment scorable, the sentence count is floored at one — paste a title with no full stop and it still returns a score rather than dividing by zero.
Syllables come from the well-worn vowel-group heuristic, not a dictionary. Each word is lowercased and stripped to letters; any word of three letters or fewer counts as one syllable; otherwise a silent trailing “e”, an “-es” or “-ed” ending after a consonant, and a leading “y” are removed, then runs of one or two vowels (a, e, i, o, u, and y) are counted. The result never drops below one. It is fast and needs no lookup table, at the cost of misjudging some words.
Because it counts vowel groups, the heuristic misses silent syllables and diphthong edge cases. “Queue” reads as one group and “business” loses a beat, while names, acronyms, and loanwords are the least reliable of all. For ordinary English prose the aggregate is close enough to be useful; on a handful of unusual words it will be off, which is why the scores are a guide rather than a measurement.
Reading-ease scores are sorted into seven plain-language bands: Very Easy at 90 and above (about 5th grade), Easy from 80, Fairly Easy from 70, Standard from 60 (8th–9th grade), Fairly Difficult from 50, Difficult from 30 (college), and Very Confusing below 30 (college graduate). The top and bottom bands are open-ended, so a score above 100 or below 0 still lands in a band rather than falling off the scale.
Common mistakes
- Reading the two scores as measuring different things. Flesch Reading Ease and Flesch–Kincaid Grade are computed from the same two averages; a high reading ease and a low grade level are two views of the same text, not independent checks. If one moves, the other moves with it.
- Chasing the score on a tiny sample. On a single line the averages are volatile — one long word can swing reading ease by dozens of points. The formulas were built for passages, so score a paragraph or more; a fragment gives a noisy, over-confident number.
- Trusting the syllable count on names and jargon. The vowel-group heuristic misjudges proper nouns, acronyms, and loanwords, so a paragraph thick with them will score harder or easier than it truly reads. Sanity-check any surprising result against the sentence you actually wrote.
- Optimising the number instead of the writing. You can lift reading ease by hacking sentences into fragments, but choppy prose is not automatically clearer. Use the score to find heavy passages, then rewrite for a reader — the figure is a symptom, not the goal.
- Applying one band to every audience. “Standard” (60–70) suits general web copy, but technical documentation for specialists, or a children’s book, have entirely different targets. Decide the grade you are aiming for first, then read the score against that, not against a universal ideal.
- Forgetting which lever to pull. When a score is low, the sentence and syllable averages on screen tell you the cause: a high words-per-sentence figure calls for splitting sentences, while a high syllables-per-word figure calls for plainer vocabulary. Guessing wrong wastes an editing pass.