Readability Checker: Flesch Reading Ease and Flesch-Kincaid Score

What is a readability score?

A readability score is a number computed from the surface shape of a text — how long its sentences run, how long its words are — that stands in for roughly what US school grade a reader would need to follow it comfortably. This readability checker computes six of the standard formulas at once: Flesch Reading Ease, Flesch-Kincaid Grade, Gunning Fog, SMOG, Coleman-Liau, and the Automated Readability Index, alongside the underlying statistics — word count, sentence count, syllable count — so you can see a Flesch reading ease score and a Flesch-Kincaid grade side by side instead of trusting whichever single number a random checker happened to pick.

Everything runs the moment you type. There is no upload and no account: paste an article, an email, a product description, or a school essay into the box, and the page recalculates every formula, every statistic, and the list of your longest sentences as you edit, so you can watch a score move in real time as you shorten a sentence or swap out a long word.

How to check readability

1. Paste or type your text into the box. There is one large text area for your writing. As you type or paste, every score and statistic on the page recalculates immediately — nothing is sent anywhere to make that happen.
2. Read the headline score. The panel above the table shows the Flesch Reading Ease score as a large number with its band label, from Very easy to Very difficult, and the matching Flesch-Kincaid grade level right beside it.
3. Check the full table, the statistics, and your longest sentences. Below the headline, a table lists all six formulas side by side, a statistics row shows word count, sentence count, syllable count, words per sentence, syllables per word, and the share of words with three or more syllables, and a longest-sentences panel lists your longest sentences with their word counts — the one part of the page you can act on directly.

What each formula measures

All six formulas read only two things about a text: how long its sentences are, and how long its words are. Flesch Reading Ease, Flesch-Kincaid Grade, Gunning Fog, and SMOG measure word length in syllables; Coleman-Liau and the Automated Readability Index measure it in letters instead, which is one reason a text full of short-but-multisyllabic words can score differently across the two groups. The scales point in different directions, too: Flesch Reading Ease runs from 0 to 100 upward, where a higher number means easier text, while Flesch-Kincaid Grade, Gunning Fog, SMOG, Coleman-Liau, and ARI all report a US school grade level, where a higher number means harder text.

Flesch’s original 1948 bands still set the standard for reading the Reading Ease number: 90–100 is very easy, around 5th-grade reading; 80–90 easy, 6th grade; 70–80 fairly easy, 7th grade; 60–70 standard or plain English, 8th–9th grade; 50–60 fairly difficult, 10th–12th grade; 30–50 difficult, college level; and below 30 very difficult, college-graduate level. This tool labels the headline score with its band automatically and shows the matching Flesch-Kincaid grade next to it.

Two of the six carry a warning of their own. SMOG was normed on samples of exactly 30 sentences, and its author was explicit that tables built for shorter texts are statistically invalid — most online checkers run it on a single paragraph anyway, so this tool flags the SMOG row whenever your text has fewer than 30 sentences instead of presenting the number as solid. Gunning Fog’s original definition also asks a human to exclude proper nouns, familiar jargon, and compound words from its complex-word count, and to ignore a third syllable that exists only because of an -es, -ed, or -ing ending. This tool applies the suffix rule and treats a capitalized mid-sentence word as a proper noun, but familiar jargon and compound words are judgment calls no program can make from the page alone — and that gap is the single biggest reason two Gunning Fog implementations can disagree on identical text.

Why the formulas disagree, and what a score cannot tell you

Take one real 67-word, three-sentence passage pulled from this site’s own article prose. Flesch Reading Ease scores it 61.7, in the standard band; Flesch-Kincaid Grade puts it at grade 10.2; Gunning Fog at 12.5; SMOG at 11.9; Coleman-Liau at 9.3; ARI at 10.9. That is a 3.2-grade spread — roughly high-school sophomore to college sophomore — for six formulas scoring the exact same paragraph, and it is not a fluke of one passage: measured across this site’s own articles, the five grade-level formulas disagree by 2.8 grades on average and by as much as 4.2 grades on the hardest passage measured. There is no single readability score, only several formulas that read the same sentence-length and word-length signals and land in different places, which is why this tool shows all six instead of picking one and hiding the disagreement.

Part of that disagreement traces back to something more basic: English spelling does not reliably determine pronunciation, so counting syllables from letters alone is inherently approximate. “Alone” and “abalone” end identically but are two syllables and four, and no spelling rule gets both right. Checked against the CMU Pronouncing Dictionary’s 134,000 words — where counting stressed vowel sounds gives an exact syllable count — letter rules alone are exact on 88.25% of 117,482 words and within one syllable on 99.53% of them. This tool closes most of that gap with a 1,050-word exception table taken directly from the CMU dictionary, which brings everyday vocabulary to 99.84% exact. What matters more is what the remaining error does to the score itself: across 54 real passages, scoring with this counter instead of CMU ground truth moves the Flesch score by 0.25 points on average and by up to 2.56 points in the worst case. A Flesch score is meaningful to about half a point, and a tool that prints it to two decimal places is showing you false precision.

Even a perfectly counted score has a hard ceiling on what it can tell you, because sentence length and word length are the only two inputs. It cannot tell whether the text makes sense: short words in short sentences score well even if the writing is incoherent, badly organized, factually wrong, or randomly shuffled, because none of that is measured. Because those are the only levers, a score can be gamed by chopping sentences at commas without changing a single word, which improves the number while making the prose worse to read. Technical vocabulary is punished regardless of audience — “photosynthesis” scores as hard for a biology paper written for biologists, where it is the correct and clearest word available — and proper nouns, acronyms, and numerals all get counted as ordinary words with syllables, so a text dense in names or figures scores harder than it actually reads. The grade levels are US school grades, and mapping them onto another country’s education system is approximate at best. These formulas are also English-specific: this page scores English text, while the Spanish, Portuguese, and French versions of this tool use formulas built for those languages, because running the English constants on another language produces a number that means nothing at all.

Why is it free?

This checker runs entirely in your browser: the text you paste is analyzed on your own device and never sent to a server, so there is nothing to upload and nothing about your writing that leaves your tab. That is true of every tool on FreeToGenerate.com — the site only calls a server at all for the handful of pages that need a language model, and this is not one of them.

There is no account to create, no watermark on anything you copy out, and no premium tier that unlocks a seventh formula or a longer text limit. It is free because computing a syllable count and a sentence count is cheap — your own device does the arithmetic instantly, and shipping the tool costs less than metering it would.