Content Readability Checker
Paste your draft and get the Flesch Reading Ease score, grade level, word and character counts and reading time as you type — plus every sentence over 25 words highlighted in your own text, so you can see where the problem actually is.
Your text
Where the long sentences are
Highlighted = a sentence of more than 25 words. This pane is read-only — edit in the box above and the highlights update.
Reading Ease = 206.835 − 1.015 × (words ÷ sentences) − 84.6 × (syllables ÷ words)Grade level = 0.39 × (words ÷ sentences) + 11.8 × (syllables ÷ words) − 15.59Syllable counts are estimated by a vowel-group heuristic, not looked up in a dictionary, so both numbers are approximate. Details below.
How to use it
- Paste the whole piece, not a sample. Readability is an average, so a 200-word excerpt can score very differently from the 2,000-word post it came from.
- Read the score with the grade level. They use the same two inputs, so they always move together — the grade level is just the friendlier way to say it to a client.
- Go straight to the highlighted sentences. The score says there is a problem; the highlights say which sentence to fix. Splitting the three worst offenders usually moves the number more than any other edit.
- Ignore what does not apply. If your reader is a tax consultant and your long words are tax words, a "difficult" score describes your subject, not a fault.
What the Flesch Reading Ease bands mean
These are the bands Rudolf Flesch published with the formula itself — its own documentation, not a Digital Hangover benchmark and not a target anyone has to hit.
| Score | Band | Roughly comparable to |
|---|---|---|
| 90–100 | Very easy | Short sentences, everyday words — a children's book |
| 80–89 | Easy | Conversational writing, popular fiction |
| 70–79 | Fairly easy | Consumer-facing web copy at its plainest |
| 60–69 | Standard | Most mainstream journalism and good blog writing |
| 50–59 | Fairly difficult | Trade press, considered B2B writing |
| 30–49 | Difficult | Academic and technical prose |
| Below 30 | Very difficult | Legal, medical and research writing |
The score can fall outside 0–100 in both directions, and the tool shows the real number rather than clamping it — a negative score is a useful signal that one sentence has gone badly wrong.
How the counting works
Every readability tool decides what a word and a sentence are, and most never say. Here is what this one does:
- Words are whitespace-separated tokens containing at least one letter or digit, punctuation stripped. Hyphenated compounds count as one.
- Sentences end at a full stop, question mark, exclamation mark or ellipsis followed by a space or the end of the text. Abbreviations (Mr., Dr., etc., Ltd., Rs.), single-letter initials and decimal points are not endings. A line break is, so headings and bullets count as their own sentence.
- Paragraphs are non-empty lines, so single- and double-spaced pastes give the same answer.
- Syllables are estimated: consecutive vowels count as one group, a silent trailing "e" is dropped, "-ed" is dropped unless it follows a t or a d, a consonant + "-le" adds one back, and any word of three letters or fewer counts as one. It is an approximation of English pronunciation, not a dictionary lookup, so words like "business" and "created" come out wrong by one.
- Reading time uses 200 and 250 words per minute — both assumptions, shown side by side because published estimates of adult reading speed vary widely.
The passive-voice and adverb checks
The passive check is a pattern match, and we would rather name the pattern than call it intelligence: a form of "to be" — is, are, was, were, be, been, being — optionally an adverb, then something that looks like a past participle (a word ending "-ed", or a common irregular like done, given, taken, written, built, sold). Every match is listed under the preview so you can check it.
It over-counts — "was tired" and "is interested" are adjectives, not passives, and both get flagged — and it under-counts, because participles the list does not know slip through and passives without an auxiliary ("the report, written in haste, …") are invisible to it. Treat it as a prompt to reread, not a defect count. The adverb check is cruder still: words ending in "-ly", minus an exclusion list for the ones that are not adverbs (only, family, supply, likely, friendly). Adverbs are not a crime; a high count is just worth a look.
What this tool cannot tell you
Flesch Reading Ease and Flesch–Kincaid grade level are heuristics built for English prose. They are not a quality score. They count sentence lengths and syllables, which means they cannot see whether your writing is accurate, useful, original or well argued — and they penalise technical vocabulary your reader may understand perfectly well. A low score on a page written for specialists is not automatically a problem, and a very high score can just mean thin writing.
They were also built for continuous English prose, so pricing tables, code samples and heavily bulleted pages produce a number but not a meaningful one. Syllable counts are estimated, so both Flesch figures are approximate; the passive and adverb checks both over- and under-count. And the tool does not rewrite anything with AI — it measures, and hands the editing decisions back to you.
Where to go next
Readability is one input; the decisions upstream of it matter more. Our content marketing strategy guide covers what to publish and why, and these content marketing examples show it in finished work. For the search side, the SEO guide is the full picture, and the keyword density checker is the next tool once the sentences read cleanly. The rest of our free tools are here.
Frequently asked questions
What is a good readability score?
It depends entirely on who is reading. Flesch's own bands put 60–69 at "standard", which is where most mainstream journalism sits, and a lot of consumer web copy aims a little higher than that. But a specialist B2B page scoring in the forties is not broken — it is written in the vocabulary its reader uses every day. Pick a band that matches your audience and use the score to check you have not drifted, rather than treating any number as a pass mark.
How accurate is the syllable count?
Close, but not exact. The tool estimates syllables with a vowel-group heuristic — consecutive vowels as one group, silent trailing "e" removed, "-ed" removed unless it follows a t or a d, consonant + "-le" adding one back — instead of looking words up in a pronunciation dictionary. It handles ordinary English well and gets a minority of words wrong by one syllable, which moves the Flesch score slightly. It is the same class of approximation almost every free readability tool uses; we just say so.
Why does the passive-voice count flag sentences that are not passive?
Because it is a pattern match, not a parser. It looks for "to be" plus something that looks like a past participle, which correctly catches "the report was written by the team" and incorrectly catches "she was tired". Every match is listed under the preview so you can dismiss the false ones in a couple of seconds. The count is there to prompt a reread, not to be a score.
Does a higher readability score help SEO?
Not directly — Google has never confirmed a readability formula as a ranking factor, and pages written for specialists routinely outrank simpler pages on their own terms. What readability affects is whether the reader stays and understands what you offered. Write for the reader, not for the score.
Is my text sent anywhere when I use this?
No. The whole analysis runs in your browser in JavaScript. Nothing you paste is transmitted to us, logged, or stored in your browser — close the tab and it is gone. That is also why there is no word limit: we are not paying for the compute, your laptop is.
The score is fine. Is the content working?
Send us a page you are not happy with. We will tell you honestly whether the problem is the writing, the structure, the brief behind it — or the fact that nobody is looking for it.