Readability Checker

What a readability score actually measures

Every formula on this page is arithmetic over three counts: how many words, how many sentences, and how long the words are - measured either in syllables or in letters. None of them reads the text. That is worth holding on to, because it explains both what the scores are good for and where they mislead.

What they capture well is mechanical load. Long sentences make a reader hold more in memory before the clause resolves; long words are more often the technical, abstract or Latinate ones. Both are real costs, and both are measurable without understanding a single sentence.

What they cannot capture is whether the writing is any good. A page of short sentences made of short words, in no sensible order, scores beautifully. So does a paragraph that defines a term three pages after first using it. If the score is the only thing you check, you will optimise for the wrong thing.

Reading the six against each other

The formulas split into two families, and the split is the useful part.

Flesch Reading Ease, Flesch-Kincaid, Gunning Fog and SMOG all count syllables. They punish long words hard. Coleman-Liau and ARI count letters per word instead, which sidesteps syllable-counting error entirely - useful when your text is full of names, acronyms or unusual spellings that a syllable heuristic will get wrong.

So when Gunning Fog reads several years above Coleman-Liau, your words are long relative to their letter count - a lot of polysyllabic vocabulary. When the two agree but both are high, the problem is sentence length instead. That diagnosis is more actionable than any single number.

• Flesch Reading Ease - 0 to 100, higher is easier. Academic prose typically 30-50.
• Flesch-Kincaid Grade - the same inputs on a US school-year scale. The safest default.
• Gunning Fog - sentence length plus the share of 3+ syllable words.
• SMOG - built only on long-word counts; the standard for health materials. Wants 30 sentences.
• Coleman-Liau - letters per word, not syllables. Steadier on names and acronyms.
• Automated Readability Index - also letter-based, built for real-time scoring.

How to bring a score down without gutting the writing

Cut sentence length first. It is the input every formula shares, it is usually where the load actually is, and it costs nothing: a 40-word sentence carrying three ideas becomes three sentences carrying one each, with no vocabulary lost.

Then look at the long words the tool counted - the share is shown above. Some are load-bearing terms you must keep, and keeping them is the right call. Others are ordinary words wearing a longer coat: "utilise" for "use", "necessitate" for "need", "in order to" for "to". Those cost you readability and buy nothing.

Do not chase a target number. A supervisor who asks for "grade 12" is asking for prose a competent reader can follow, not for a specific output from a 1975 formula. Use the score to find the two worst paragraphs, fix those, and stop.

Where these formulas came from

Flesch Reading Ease dates from 1948; Flesch-Kincaid was fitted in 1975 on US Navy training manuals for enlisted personnel. Gunning Fog came out of business writing in 1952, SMOG from health communication in 1969, and Coleman-Liau and ARI were designed for machine scoring, which is why both avoid syllables.

Not one of them was validated on academic prose. Applied to a paper they read systematically harder than the audience finds it, because a specialist reader knows the long words and the formula cannot know that. This is a real limitation and not a small one - treat the absolute number as approximate and the changes between drafts as the signal.

Common questions