Paste any writing. It scores how hard each sentence is to read — reading ease, school grade, and a shade behind every sentence — entirely inside this page. Nothing is uploaded.
—
Both formulas need a syllable count, and there is no exact rule for English — only guesswork over vowel groups. Here is the guess this page makes, step by step, plus every word it is known to get wrong.
Two formulas from the 1940s–70s, both built from the same three counts: words, sentences, syllables.
That is the whole of it. Neither formula reads anything. They count length and nothing else, so a sentence of short familiar words scores well even if it is nonsense, and a short sentence of long precise words scores badly even if it is perfectly clear. Treat the number as a smoke alarm, not a judge.
Reading ease is meant to land between 0 and 100, but the formula has no ceiling or floor: very short plain sentences overshoot 100, and very long ones go negative. The dials clamp; the numbers do not.
Syllables are estimated, not counted. The rule: lowercase the word, drop a silent
-e / -es / -ed ending, drop a leading y, then count
runs of a e i o u y. One patch on top: a word ending in consonant + vowel + -ing
(be·ing, go·ing) gets one more. A word containing digits is scored as
one syllable per digit. Against a built-in list of hand-counted words the rule is right about four times in
five — the exact score and the full list of failures are in the panel above. Words on that list use the
hand count; everything else uses the rule.
Sentence splitting is a guess too. The page breaks on . ! ?
followed by whitespace, then unbreaks the obvious traps: common abbreviations, single-letter initials,
decimals, and a full stop followed by a lowercase word. Semicolons do not split, so a semicolon-spliced
sentence is scored as the one long sentence it really is.
Everything here runs in this page: no network requests, no storage, no analytics. Your text stays in the tab.