Estimate the U.S. grade level needed to understand the text using character and word counts. Fast and reliable.
Built and maintained by Paul Clark, Redmoon Software · Last verified · Sources
Use Coleman–Liau when you need a robust automated score — it has no syllable counter, so it cannot be tripped up by unusual words or foreign names. It is the safest choice for OCR’d text, social-media posts, or content with lots of proper nouns.
Coleman–Liau and ARI both rely on character counts and typically produce similar grade levels. Compared to Flesch–Kincaid, they are more robust to unusual vocabulary but blind to word familiarity. Use Coleman–Liau when automating; use FK when communicating with humans who already know it.
The Coleman-Liau index was designed in 1975 specifically for automated computing — it requires no syllable counting.
It uses characters per word and sentences per 100 words to estimate US grade level.
Because it only needs character counts, it generalises better across languages with regular spelling.
Characters are easier to count programmatically and don't require language-specific syllable rules. This makes Coleman-Liau a popular choice for automated tools.
Yes — Coleman-Liau correlates well with cloze test scores, the gold standard of readability research.
L is the average number of letters per 100 words and S is the average number of sentences per 100 words. The result is a U.S. grade level, so a score of 9 means the text suits a ninth-grade reader.
Coleman-Liau judges word difficulty by letter count while Flesch-Kincaid uses syllables, so a text full of short but multi-syllable words can score differently. For most prose the two land within a grade or two of each other.
Input
Coleman–Liau needs no syllable count. It uses letters and sentences.
Output
Coleman–Liau Index: 8.5 — 8th–9th grade.
About 4.6 letters per word and ~7 words per sentence. The character-based formula handles the punctuation and lengths directly, no syllable counting required.
Coleman–Liau estimates a grade level from characters per word and sentences per 100 words. It never counts a syllable, and that was the design goal in 1975: the authors wanted a formula a machine could compute reliably from raw text, at a time when syllable counting meant either a dictionary lookup or a heuristic that would be wrong some of the time.
Fifty years later that decision still matters here, because this tool's syllable counter is exactly the kind of heuristic they were avoiding. Coleman–Liau has no such error term. On text heavy with proper nouns, technical terms or unusual spellings — where a syllable estimator is least reliable — it is the more trustworthy of the two approaches.
The trade-off is that character length is only a proxy for difficulty, and the proxy breaks in both directions. Short difficult words ("writ", "tort", "bond" in its financial sense) score as easy. Long familiar words ("grandmother", "understanding", "immediately") score as hard.
It also generalises across languages better than the syllable-based formulas — the character-length relationship holds reasonably in languages with regular spelling — but the grade-level constants remain calibrated on English schooling, so treat a non-English result as a relative measure only.
Practical use: run Coleman–Liau alongside a syllable-based score. Where they agree you can be confident. Where they diverge sharply, the vocabulary is doing something unusual, and that is worth looking at directly rather than resolving with a number.
The method behind this tool, so you can check it rather than take it on trust.
Coleman, M. & Liau, T.L. (1975). “A computer readability formula designed for machine scoring.” Journal of Applied Psychology, 60(2), 283–284.
Deliberately uses characters rather than syllables, so it can be computed without a pronunciation dictionary.
Classic 0–100 readability score.
Readability scoresU.S. grade level for your text.
Readability scoresPenalises complex (3+ syllable) words.
Readability scoresHealth & medical text grade level.