Text tools / English prose
Readability Checker
See how sentence length and word structure shape three readability estimates. Inspect the counts, review passages and compare your next revision.
Ready for English text. Analysis runs when you select Analyze text.
Revision baseline
Your readability estimates
These formulas describe surface features of English text. They do not measure comprehension, factual accuracy, writing quality, readers' ability or search ranking.
Flesch Reading Ease
Flesch-Kincaid
Estimated US grade scale. Lower values reflect shorter sentences or fewer syllables per word.
Coleman-Liau
Estimated US grade scale using letters and sentence length. No syllable count needed.
Counts behind the estimates
Places to review
Read these passages in context. A long sentence or a familiar long word can be exactly right for your audience.
Long sentences
Words estimated at 3+ syllables
Use the results to guide a revision
- Paste a complete passage of English prose, ideally at least 100 words across several sentences. Remove navigation, code, links and markup.
- Analyze and check the sentence count. For a list, try the each-line option. Review the long sentences and confirm that the estimated syllables fit your vocabulary.
- Select Use as baseline, edit your draft, and analyze again. The comparison keeps aggregate counts in this tab. It does not retain a backup of your original text.
- Read the revision aloud and check it with intended readers. Keep essential terms and meaning. Copy or download a report when you want to keep the results.
What the formulas mean
Higher Flesch Reading Ease usually reflects shorter words and sentences. The familiar range is 0 to 100, but the equation can produce values outside it. Flesch-Kincaid and Coleman-Liau use historical US grade scales; a decimal is an estimate, not an exact school placement or a judgment about anyone's education. We show the raw results to one decimal place and do not average them.
Formulas and a worked example
W = words; S = estimated sentences; Y = syllables used; L = letters.
Reading Ease = 206.835 - 1.015 × (W / S) - 84.6 × (Y / W)
Flesch-Kincaid = 0.39 × (W / S) + 11.8 × (Y / W) - 15.59
Coleman-Liau = 0.0588 × (100 × L / W) - 0.296 × (100 × S / W) - 15.8
For 100 words, 5 sentences, 150 syllables and 450 letters, the results are 59.6 Reading Ease, 9.9 Flesch-Kincaid and 9.2 Coleman-Liau. Rounding happens only at display time.
References: Flesch's original article, reprinted by ERIC (PDF), Kincaid and colleagues' report record, Microsoft's formula documentation, and Harvard's Coleman-Liau specification.
Exactly how this tool counts
Words are runs of Latin letters or digits. Internal apostrophes and hyphens keep contractions and compounds together. Decimal and comma-separated numeric runs count as one word. Letters count without punctuation, digits or combining accent marks. Emoji and other symbols are ignored. Unicode accents and curly apostrophes are supported; text is normalized to NFC. The character limit uses browser text units, so an emoji can use two units.
Periods, question marks and exclamation marks end estimated sentences. Blank lines also end them; the each-line option adds every line break. A final fragment counts as a sentence. Repeated terminal punctuation counts once. Common titles, initials, dotted abbreviations and decimal points are protected, but an abbreviation at the end of a sentence may still merge two sentences. Semicolons and colons do not split sentences. These are deterministic heuristics, not grammatical parsing.
Syllables use English vowel groups, common suffix rules and a small exception list. Hyphenated parts are added together. Uppercase abbreviations use estimated letter names; all-digit tokens use one placeholder syllable. Numbers, names, accents and technical vocabulary need checking. A corrected total changes Reading Ease and Flesch-Kincaid only; the word-review list still uses automatic per-word estimates.
Limits and privacy
Use this checker for English prose. It cannot detect whether Latin-script text is English and rejects non-Latin letters. A very short sample, fragmented list or unusual punctuation can move scores sharply. Below 100 words or 3 sentences, a caution is shown; this is a practical warning threshold, not a statistical confidence interval. Even longer passages need human review.
These formulas omit meaning, prior knowledge, layout, grammar and the reader's goals. They cannot prove understanding, accessibility compliance or search performance. Review the actual text with the people who will use it.
The tool makes no network requests and saves no input to browser storage. Text, reports and the baseline exist in this page's memory until you clear or leave it. Copying and downloading happen only when you select those actions. Reports include review excerpts. The site's shared page services, including any centrally configured analytics, are separate from this calculation.