Readability Score Analyzer
Paste any text to compute six classic readability scores at once — Flesch Reading Ease, Flesch-Kincaid Grade, Gunning Fog, SMOG, Coleman-Liau and Automated Readability Index — plus an audience grade level and a hard-sentence heatmap. Runs entirely in your browser.
What is a readability score?
A readability score estimates how easy a piece of text is to read, based on objective signals like average sentence length, syllable count, character count and the proportion of long words. Each formula returns a number that can be mapped to a U.S. school grade level or to a difficulty band. Lower grade levels mean the text is accessible to a wider audience; higher grade levels mean the reader needs more formal education to follow comfortably.
This readability analyzer runs the six most widely used formulas in parallel — Flesch Reading Ease, Flesch-Kincaid Grade, Gunning Fog Index, SMOG, Coleman-Liau and the Automated Readability Index — so you don't have to choose one. The tool also returns a recommended audience (elementary, middle school, high school, college, professional) and highlights the sentences slowing your reader down in a color-coded heatmap.
How to use the readability analyzer
- Paste your text in the editor — an essay, blog post, email, press release, instruction manual or product description. Everything is processed locally; nothing is uploaded.
- Read the summary card on the right: it shows the recommended audience and the median grade level across all formulas. This is the single number to remember.
- Compare the six scores. If they agree within ~1 grade, the verdict is reliable. If they disagree by 3+ grades, your text mixes simple and complex passages — inspect the heatmap.
- Open the heatmap at the bottom. Hard and very hard sentences are highlighted. Rewrite or split those first — it usually moves the average grade down by 2-4 levels with little effort.
- Re-run live. The analyzer recomputes every keystroke, so you can iterate until you hit the target audience.
The six readability formulas explained
Flesch Reading Ease
The oldest and most cited score (Flesch, 1948). It runs on a 0–100 scale where higher is easier: 90–100 is understandable by a fifth-grader, 60–70 by an average adult, below 30 only by university graduates. The formula is 206.835 − 1.015 × (words / sentences) − 84.6 × (syllables / word). Microsoft Word and Google Docs surface this number in their built-in writing tools.
Flesch-Kincaid Grade Level
A 1975 reformulation of Flesch by Peter Kincaid for the U.S. Navy that returns a U.S. school grade instead of a 0–100 score. Grade 8 means a typical 13–14-year-old can read the text. The U.S. Department of Defense and several state governments mandate Flesch-Kincaid Grade ≤ 8 for consumer-facing documents. The formula is 0.39 × (words / sentences) + 11.8 × (syllables / word) − 15.59.
Gunning Fog Index
Robert Gunning's 1952 index weighs long words (3+ syllables) more heavily than short ones. The number is the years of formal education needed: Fog 8 = 8th grade, Fog 12 = high-school senior, Fog 17 = college graduate. Newsroom style guides (Reuters, AP) target Fog 8–11 for general-audience news. The formula is 0.4 × [ (words / sentences) + 100 × (complex words / words) ].
SMOG Index
The Simple Measure of Gobbledygook (McLaughlin, 1969) was designed for healthcare materials and is the formula the U.S. National Institutes of Health recommends for patient information. It estimates the years of education needed to fully understand the text (versus only partly, which is what Flesch-Kincaid measures). SMOG is statistically most accurate on samples of 30 sentences or more. The formula is 1.0430 × √(polysyllables × 30 / sentences) + 3.1291.
Coleman-Liau Index
Meri Coleman and T. L. Liau (1975) noticed that character counting is more reliable than syllable counting for automated systems, since algorithms cannot perfectly detect syllables in proper names, acronyms and loanwords. Coleman-Liau therefore uses letters and sentences only: 0.0588 × L − 0.296 × S − 15.8, where L is letters per 100 words and S is sentences per 100 words. The output is a U.S. grade level.
Automated Readability Index (ARI)
The U.S. Air Force's 1967 formula, also character-based and grade-leveled: 4.71 × (characters / words) + 0.5 × (words / sentences) − 21.43. ARI tends to give slightly lower grades than Flesch-Kincaid on the same text because long single-syllable words like "strengths" inflate syllable-based metrics but barely move ARI. Useful as a second opinion alongside Flesch-Kincaid.
Reading levels and their audiences
The grade-level scores can be mapped to typical audiences. Pick the row that matches your target reader and aim for that range across all six formulas.
| Audience | Grade range | Flesch Reading Ease |
|---|---|---|
| Elementary school (8–11 yrs) | 1–5 | 90–100 |
| Middle school (11–14 yrs) | 6–8 | 70–89 |
| High school (14–18 yrs) | 9–12 | 50–69 |
| College undergraduate | 13–15 | 30–49 |
| College graduate / professional | 16+ | 0–29 |
Recommended targets by content type
- Marketing landing pages — Flesch 60–70 / FK Grade 7–8. Most U.S. adults read at this level.
- Blog posts & news — Flesch 50–60 / FK Grade 8–10. Reuters and the BBC target this band.
- Healthcare patient information — SMOG ≤ 6 / FK Grade ≤ 6. NIH and AMA recommendation.
- Government & legal documents for the public — FK Grade ≤ 8. U.S. Plain Writing Act of 2010.
- Technical documentation (developers) — FK Grade 10–13 is acceptable; the audience is specialised.
- Academic papers — FK Grade 14+ is normal; readability is not the goal, precision is.
- Children's books (age 6–8) — Flesch 90+ / FK Grade ≤ 3.
Tips to improve a poor readability score
The two levers that move every formula at once are sentence length and word length. Quick fixes that consistently work:
- Split long sentences. Anything over 25 words is a candidate. Look for "and", "but", "because", semicolons — they often mark natural break points.
- Replace long words with short ones: "utilize" → "use", "approximately" → "about", "in order to" → "to", "facilitate" → "help".
- Drop unnecessary qualifiers: "very", "really", "quite", "rather". They add length without meaning.
- Use active voice. Active sentences are shorter and easier to parse than passive ones.
- One idea per sentence. If a sentence contains two facts, make it two sentences.
- Re-run the analyzer after each pass. Most texts move 2–4 grade levels with a single rewrite cycle.
Frequently Asked Questions
All six are reliable for English prose. When in doubt, trust the median grade level shown in the summary card — it averages out the quirks of individual formulas. If your scores disagree by more than three grades, your text is uneven and the heatmap will show why.
The analyzer uses a heuristic algorithm that counts vowel groups, with adjustments for silent -e, -es, -ed and short words. It is the same approach used by Microsoft Word's readability statistics and matches dictionary syllable counts on roughly 95% of common English words. Unusual proper names and loanwords can be off by one syllable.
Words with three or more syllables. SMOG counts them directly; Gunning Fog calls them "complex words" and includes them in its formula. Common suffixes like -es, -ed and -ing are not counted as extra syllables in the strictest interpretation, but most online tools (including this one) treat any 3+ syllable word as complex.
The SMOG formula was statistically validated on samples of 30 sentences or more. Below that, the result becomes unreliable because a single polysyllabic word in a 5-sentence sample skews the index disproportionately. The tool still computes SMOG for short texts but you should weight it less than the other five.
The formulas were calibrated on English. They still run on French, Spanish, German or Italian text, but the grade level is not comparable to the original English scale. For other languages, look at the relative score before and after editing rather than at the absolute number. For Chinese, Japanese, Thai, Arabic and similar scripts the syllable heuristic is unreliable — use the character count from the meta stats row instead.
No. All computation runs locally in your browser using JavaScript. The text never leaves your device, and the tool works offline once the page is loaded.
For general U.S. audiences, target Flesch-Kincaid Grade 7–8 (matches the average adult reading level reported by the National Assessment of Adult Literacy). For healthcare, social services and government documents intended for vulnerable populations, target Grade 5–6. For technical or academic readers, Grade 10–14 is acceptable.
Yes — and they often reveal one of the most common AI tells: a uniformly high grade level (typically 12–14) with very few short sentences. Human writing tends to alternate short and long sentences, which lowers the average grade. The heatmap will show this pattern at a glance: AI text usually has almost no green sentences.