Five standard formulas, the raw counts behind them, and a plain-language reading of what the numbers mean - plus the CEFR band, which readability formulas alone will not give you.
Readability formulas all do the same basic thing: they count how long your sentences are and how long your words are, and turn that into a school grade. They were built in the 1940s to 1970s for military manuals, insurance documents and health information, and they are still the fastest way to catch writing that has drifted out of reach of its audience.
This tool runs the four grade-scale formulas that appear in most style guides - Flesch-Kincaid Grade, Gunning Fog, SMOG and the Automated Readability Index - alongside the Flesch Reading Ease score, and reports the average of the four grade scores as a consensus. It also shows you the counts each formula consumed, so you can see whether a high score came from long sentences or from long words. Those two problems have different fixes.
This checker also estimates a CEFR band. A US grade level is the wrong unit if your reader is a Turkish student learning English rather than an American twelve-year-old, and the two scales do not map onto each other cleanly.
0 words
Paste or type some text to see its scores.
Everything runs in your browser. Your text never leaves this page, is never uploaded, and is never stored.
Which formula should you use?
Flesch Reading Ease returns a 0-100 score where higher is easier; 60-70 is considered standard English. It is the formula most style guides and content platforms quote, and it is the one to use when you want a single number that non-specialists understand.
Flesch-Kincaid Grade converts the same inputs to a US school grade, which makes it easier to act on: grade 8 means an eighth-grader could read it. It is written into US federal plain-language guidance and into many insurance regulations.
Gunning Fog counts words of three or more syllables rather than syllables per word, which makes it harsher on nominalised business writing - the "implementation of the utilisation" register. If your text scores much worse on Fog than on Flesch-Kincaid, you have a long-word problem rather than a long-sentence problem.
SMOG was designed for health-care materials, where the cost of a reader misunderstanding is high, and it targets 100% comprehension rather than the 50-75% the Flesch formulas assume. It therefore reads about two grades harsher. ARI uses characters instead of syllables, which makes it the most mechanically reliable of the four, since character counting has no edge cases while syllable counting has hundreds.
How to bring a score down without dumbing the text down
Split the long sentences first. Sentence length is an input to all four formulas, so it is the single highest-leverage change, and the 90th-percentile figure in the counts above tells you whether you have a few very long sentences hiding behind a lot of short ones. Cutting a 45-word sentence into two 22-word sentences moves every score at once.
Then look at the three-syllable words. Most of them in business and academic writing are nominalisations - "make an assessment of" instead of "assess", "provide assistance to" instead of "help". Undoing those shortens the sentence and the words together.
What you should not do is chase the number. These formulas cannot see whether a sentence is ambiguous, whether the paragraph order makes sense, or whether the reader has the background knowledge the text assumes. A text can score grade 6 and still be incomprehensible. Treat the score as a smoke alarm, not as an editor.
Why a grade level is the wrong unit for language learners
Grade-level formulas were calibrated on native-speaking American schoolchildren. Those children have a spoken vocabulary of thousands of words before they read a single sentence, so the formulas can assume that any common word is known and only measure structural load.
That assumption fails for a language learner. A B1 student may handle a long sentence comfortably while being stopped dead by an everyday word they have simply never met. This is why a text can score grade 7 and still be a C1 text: the sentences are short, but the vocabulary sits outside the first 5,000 words.
The CEFR band shown alongside the formulas is computed from a frequency-banded wordlist rather than from syllable counts, so it catches exactly that case. If you are choosing material for an English class, it is the number to trust; if you are editing a public-information leaflet for native speakers, the grade level is.
Reading practice that already sits at the right level
If you are levelling material for learners, the hard part is not measuring a text - it is finding one that lands where you need it. LingoRead generates reading passages at a CEFR level you choose, on a topic you choose, with vocabulary lists and comprehension questions attached.