How to Check a Readability Score Online: Seven Formulas Explained in Plain English

Direct answer: You can check a readability score online with the Readability Checker tool: paste your text and it scores it live as you type against seven established formulas — Flesch Reading Ease, Flesch-Kincaid Grade, Gunning Fog, SMOG, Coleman-Liau, Automated Readability Index, and Dale-Chall — while a problem finder lists sentences over 20 words and the longest words in the passage. Nothing is uploaded; the scoring happens in your browser.

A readability score is a rough measurement of how easy a passage is to read, based on countable features such as sentence length, word length, and syllable counts. It is not a judgement of quality, accuracy, or style. A dense legal clause can score as difficult while being perfectly correct, and a simple sentence can score as easy while being wrong. The score answers one narrow question — how much effort does decoding this text demand — and that narrow answer is genuinely useful when you are writing for a general audience.

The reason to check a score is practical, not academic. Public-facing pages, help articles, product descriptions, and school materials all serve readers who did not choose to wrestle with the text. If the score says a page reads at a high grade level, some of those readers will leave rather than struggle. A quick check before publishing catches that problem while it is still cheap to fix, by shortening a few sentences and swapping a few long words.

The seven formulas at a glance

The seven formulas share an idea — longer sentences and longer words make text harder — but they weigh the signals differently. Flesch Reading Ease gives a score from 0 to 100, where higher means easier; a score of 60 to 70 is standard plain English, and 70 or above reads comfortably for most adults. Flesch-Kincaid Grade converts a similar calculation into a US school grade, so a grade of 8 means a typical eighth-grader could read it. Gunning Fog and SMOG also produce grade levels, with SMOG leaning heavily on multi-syllable words. Coleman-Liau uses characters rather than syllables, which makes it steadier for texts with unusual words, and the Automated Readability Index uses characters and words per sentence to estimate a grade. Dale-Chall compares the vocabulary against a list of familiar words, so it penalises unfamiliar vocabulary even when the words are short.

Seeing all seven together is more informative than seeing any one alone, because their disagreements are themselves information. A passage with short sentences but technical vocabulary may score well on sentence-based formulas and poorly on Dale-Chall. A passage with long sentences but simple words may show the opposite pattern. When the formulas agree, you can trust the verdict; when they spread out, the spread tells you which kind of difficulty your text has — sentence difficulty, word difficulty, or vocabulary difficulty — and that tells you what to fix.

Public pages commonly aim for grade 8 or below on the grade-level formulas, with a Flesch score in the 60 to 70 range or higher. That is a guideline for general audiences, not a rule for every document: a medical leaflet, a research summary, or a specialist manual may legitimately read harder, and forcing it down to grade 8 by chopping necessary terms out would harm accuracy. Use the target that fits your readers, and be honest about who those readers are.

The Flesch formula, worked through with real numbers

Flesch Reading Ease is the best formula to understand first, because its arithmetic is open and its result is intuitive. The formula is 206.835 − 1.015 × (words divided by sentences) − 84.6 × (syllables divided by words). In plain terms: start from 206.835, subtract a penalty for long sentences, and subtract a larger penalty for long words measured in syllables. Short sentences and short words keep the score high; long sentences and syllable-heavy words pull it down.

Take a short passage of 100 words, 8 sentences, and 150 syllables. Words per sentence come to 12.5, and syllables per word come to 1.5. The sentence penalty is 1.015 × 12.5 = 12.6875, and the word penalty is 84.6 × 1.5 = 126.9. Subtracting both from 206.835 gives 206.835 − 12.6875 − 126.9 = 67.2475, or about 67. That sits inside the 60 to 70 band described as standard plain English — a passage most adults can read comfortably, which is exactly where a general-audience page wants to be.

Now imagine the same 100 words stretched into 4 sentences with 180 syllables. Words per sentence jump to 25, syllables per word to 1.8, and the score falls to 206.835 − 25.375 − 152.28 = about 29, which reads as very difficult. Nothing about the content changed — only the sentence and word lengths — yet the reading effort roughly doubled in the formula's eyes. That is the whole lesson of readability in one example: how you package ideas matters as much as which ideas you include.

Live scoring and the problem finder

Scores update live as you type, which changes how editing feels. Instead of writing a whole draft and then checking it, you can watch the score move as you shorten a sentence or simplify a paragraph, and learn which edits actually help. That feedback loop is the tool's most useful feature for writers who are still building an instinct for plain English: a few sessions of watching scores respond to edits teaches the pattern faster than any style guide.

The problem finder turns a single number into actionable detail. It lists sentences over 20 words, which are the most likely places where a reader loses the thread, and it lists the longest words in the passage, which are the most likely places where vocabulary is doing the damage. A long sentence is not automatically bad — a well-built long sentence can carry a complex idea elegantly — but a list of them shows you exactly where to look, and most drafts improve when their three or four longest sentences are split.

Use the problem finder as a triage list, not a hit list. Work through the longest sentences first, splitting each into two where the split does not break the logic, and then look at the longest words and ask whether a shorter word says the same thing. Keep technical terms that your readers need — a medical or legal term that is the correct word should stay even if it is long — and simplify the words around it instead. Readability improves most when the simple words carry the structure and the necessary hard words stand out clearly.

The statistics behind the scores

Alongside the seven scores, the tool reports the raw counts the formulas use: words, sentences, syllables, characters, and complex words, defined as words with three or more syllables. Those counts are worth glancing at, because they explain the scores: a high complex-word count with a low sentence count points at vocabulary, while a high sentence length with a modest complex-word count points at sentence structure. Knowing which count is driving the score tells you which editing move to make.

Reading time and speaking time translate the counts into something a planner can use. Reading time assumes 200 words per minute, the usual planning figure for silent reading, while speaking time assumes 130 words per minute for narration or presentation. A 1,000-word help article therefore reads in about five minutes and takes about eight minutes to speak aloud — a useful check when a script or a voiceover is being timed, and a reality check when a page claims to be a quick read but runs to several thousand words.

Syllable counting deserves a note of honesty. The tool page documents its syllable heuristic, because counting syllables by rule is approximate: English spelling is irregular, and any automatic counter will miscount some words. That approximation is acceptable for a readability estimate, where the pattern across hundreds of words matters more than any single word's count, but it is another reason to read scores as a range rather than a precise verdict. A score of 67 and a score of 69 are the same story told with slightly different rounding.

When scores disagree, read them as a range

Scores disagree because formulas weigh different signals, and disagreement is normal rather than a sign of error. A passage can read at grade 7 on one formula and grade 11 on another without anything being wrong: the formulas are measuring different aspects of difficulty, and a text that is easy in one aspect and hard in another will legitimately score both ways. The honest reading is the range — this passage reads somewhere between grade 7 and grade 11 depending on which difficulty you count — and the editing question becomes which end of the range you want to move.

That range reading also protects against gaming the score. It is possible to push a score up by chopping sentences mechanically while leaving the ideas just as tangled, and the result reads worse even as it scores better. The score is a diagnostic, not a target to be hit by any means: edit for the reader, check the score to confirm the edit helped, and stop when the passage reads clearly rather than when a number crosses a threshold.

A final practical note: your text never leaves your browser during checking, so drafts, client material, and unpublished work can be scored without sending them anywhere. Pair the checker with a draft review in the Slug Generator when a page title needs a clean URL, or with the Keyword Density Checker when a draft's repeated phrases need a look. More writing guides live on the Asad Webs blog.

Frequently Asked Questions