Readability Checker
Paste your text below to get live Flesch Reading Ease and Flesch-Kincaid Grade Level scores.
About the Readability Checker
Flesch Reading Ease and Flesch-Kincaid Grade Level are calculated straight from your word, sentence, and syllable counts using the standard published formulas. Syllables are estimated with a vowel-group heuristic rather than a dictionary lookup, which is how most readability tools actually work, so treat the numbers as a close approximation rather than an exact count.
Two formulas measuring the same thing on different scales
Reading Ease was published by Rudolf Flesch in 1948 as a way to score text from 0 to 100, with higher numbers meaning easier to read: 206.835 - 1.015 × (words / sentences) - 84.6 × (syllables / words). Flesch-Kincaid Grade Level came later, developed in the 1970s by J. Peter Kincaid for the U.S. Navy to translate that same underlying idea into a more directly actionable number, a rough U.S. school grade level: 0.39 × (words / sentences) + 11.8 × (syllables / words) - 15.59. Both formulas are built from exactly the same three ingredients — how many words per sentence, and how many syllables per word — just weighted and scaled to produce a different kind of output, one a relative ease score, the other a concrete grade-level estimate.
The syllable heuristic reliably undercounts one specific class of words
Syllables are estimated by counting runs of consecutive vowel letters (a, e, i, o, u, y) in a word, then subtracting one if the word ends in a silent "e" and more than one vowel group was found. That correction works well for words like "large" or "like," where the trailing e genuinely isn't pronounced. But it also fires, incorrectly, on words ending in a consonant plus "le" where that final syllable actually is pronounced as its own beat — "table," "little," and "purple" all get counted as one syllable fewer than they actually have, because the heuristic can't distinguish a truly silent trailing e from the syllabic "-le" ending common in English. This is a known, systematic blind spot shared by essentially every vowel-group syllable estimator, not a bug unique to this implementation, but it's worth knowing your syllable and score numbers will run slightly low on any passage using a lot of "-le"-ending words.
Y counts as a vowel here, unlike this site's Pig Latin translator
The vowel-group pattern used for syllable counting is [aeiouy]+, deliberately including y, because treating y as a consonant would badly undercount words like "rhythm," "gym," "happy," or "sky" where it's the only vowel sound present. That's a different rule than the one on this site's Pig Latin Translator, which never treats y as a vowel under any circumstance — a reminder that "is y a vowel" doesn't have one universal answer in English; it depends entirely on what the rule is actually being used for.
Sentence detection is punctuation-only, which abbreviations and decimals can fool
Sentences are counted by matching runs of text that end in a period, exclamation point, or question mark, with no awareness of context. That means an abbreviation like "Dr." or "U.S.," or a decimal number like "3.14," gets read as ending a sentence, artificially inflating the sentence count and shortening the apparent average sentence length — which nudges both scores toward looking easier to read than the text actually is. On the other end, a chunk of text with no terminal punctuation at all (a title, a list fragment, a line of dialogue missing its period) is treated as exactly one sentence regardless of how many words it actually contains, which can just as easily skew the ratio the other way for a long, unpunctuated block.
Treat the score as unreliable on very short input
Both formulas were designed and validated against full passages of running prose, not single words or short fragments. Feeding in just a sentence or two produces a words-per-sentence and syllables-per-word ratio computed from a tiny sample, which can swing the score dramatically based on one unusually long or short word. The numbers become meaningfully stable once you're checking at least a full paragraph or two of real writing, which is also the scale these formulas were originally built and tested against.
Frequently Asked Questions
What's the difference between Flesch Reading Ease and Flesch-Kincaid Grade Level?
They're built from the exact same inputs — words per sentence and syllables per word — but scaled differently. Reading Ease produces a 0-to-100 score where higher means easier to read, while Grade Level converts that same underlying calculation into an approximate U.S. school grade level, which is often easier to act on directly.
Why do words like "table" or "little" show the wrong syllable count?
The syllable estimator counts vowel groups and subtracts one for what it assumes is a silent trailing "e." That correction is usually right, but it incorrectly fires on words ending in a consonant plus "le," where the final syllable is actually pronounced, causing those specific words to undercount by one syllable. This is a known limitation shared by most vowel-based syllable estimators, not just this one.
Does an abbreviation like "Dr." or a decimal number affect the sentence count?
Yes. Sentences are detected purely by looking for periods, exclamation points, and question marks, with no understanding of context, so an abbreviation or a decimal number gets read as a sentence ending, which can inflate the sentence count and skew the readability scores toward looking easier than the text actually is.
Is y treated as a vowel for syllable counting?
Yes, deliberately, since treating it as a consonant would badly undercount common words like "rhythm" or "happy" where y is the only vowel sound present. This is a different rule from what some other tools on this site use y for, depending on what each tool actually needs it to do.
Is the score reliable for a single sentence or a short phrase?
Not very. Both formulas were designed and validated against full passages, so a tiny sample can swing the score heavily based on just one long or short word. The numbers become meaningfully stable once you're checking at least a paragraph or two of real writing.