What a Readability Score Actually Tells You
A readability formula counts two things: how long your sentences are, and how long your words are. That is essentially it. From those two measurements it estimates how much effort a reader has to spend decoding the text before they can start thinking about what it means.
This sounds crude, and it is. But it correlates well enough with reading difficulty that the US Navy and the insurance industry (and several US states for policy readability) use these formulas to check whether their documents are usable. A score will not tell you whether your writing is any good. It will tell you whether the average reader can get through a sentence without losing the thread.
This tool runs six formulas at once, entirely in your browser. Nothing is uploaded and there is no daily limit.
How to Use the Readability Score Checker
Paste Your Text
Any written content works: an essay, a landing page, a lesson handout, a legal notice. Aim for at least 200 words so the averages settle.
Choose Your Audience
Pick who the text is for. The target grade level shifts accordingly, so a children's book is judged against grade 4 rather than grade 8.
Read the Verdict
Six scores, full statistics, flagged sentences, and specific fixes appear instantly. Export to PDF or TXT when you are done.
What Is a Good Readability Score?
There is no single perfect number. The right Flesch Reading Ease score depends entirely on who you are writing for. Use this table as a rough benchmark:
| Content Type | Target Flesch | Grade Level |
|---|---|---|
| Children's book / Early readers | 90+ | 4–5 |
| Blog post / General web copy | 60–70 | 7 |
| Business correspondence / Reports | 50–60 | 10–12 |
| Academic papers / Technical docs | 30–50 | 13+ |
Try this on an insurance policy sometime. You will usually find a section in the twenties, and the rewrite that gets it to fifty rarely removes a single fact. It just removes the words that were doing nothing. The content does not change. What changes is how many words sit between the reader and the meaning.
The Same Idea, Three Reading Levels
Here is one idea written three ways. The content barely changes. The reading difficulty changes enormously.
"Cities are hotter than the countryside. Roads and buildings soak up the sun all day. At night they let that heat back out. Trees would cool things down, but cities have fewer of them."
"Cities tend to run warmer than surrounding rural areas because asphalt and concrete absorb solar radiation during the day and release it slowly after sunset. Vegetation, which cools its surroundings through shade and evaporation, is usually scarce in dense districts."
"The urban heat island phenomenon is attributable to the differential thermal absorptivity of impervious surfaces relative to vegetated land cover, compounded by the attenuation of convective heat dissipation resulting from morphological obstruction of prevailing airflow patterns."
What to Do With a Bad Score
If your text comes back several grades above your target, these moves close the gap fastest, roughly in order of impact:
- Split your longest sentences first. Sentence length carries more weight than word choice in every formula here. Find the sentence over 35 words and cut it in two, and you will see every score move at once.
- Name the actor. Passive constructions add words without adding meaning. "The report was reviewed by the committee" becomes "The committee reviewed the report" and loses three words.
- Swap abstract nouns for verbs. "Implementation of the policy" is four syllables heavier than "implementing the policy" and six heavier than "starting the policy."
One more thing: leave the technical terms alone. If your readers need the word, keep it. Replacing "premium" with "payment" in an insurance document will improve the score and confuse the policyholder, because "premium" is the word they were given, the word on their statement, and the word they will search for. Simplifying jargon your audience already knows makes the score prettier and the writing worse.
Which Formula Should You Use?
All six formulas measure sentence and word length, but each has a slightly different bias. Here is when to trust each one:
Best for general audiences. The 0–100 scale is intuitive. Use it when you need a single, easy-to-explain number for non-experts.
Best for US education and government. It maps directly to school grades, which many plain-language mandates require.
Best for business and finance. It penalizes complex words harder than Flesch, so it catches jargon-heavy corporate writing — the place where weasel language usually hides.
Best for healthcare and public safety. McLaughlin designed it specifically for assessing health literacy. It is the most conservative — often gives the highest grade, which is exactly what you want when someone's health depends on understanding the text.
Best for OCR or transcribed text. It does not count syllables at all — it uses characters per word instead. More reliable when syllable counting might fail (e.g., scanned documents).
Character-based like Coleman-Liau, but weighted differently. Originally built for real-time scoring on typewriters, which is why it avoids syllable counting entirely. Useful as a cross-check against the Flesch family.
Where These Formulas Mislead You
Worth knowing before you optimize for a number:
- !
They cannot see meaning. "The cat sat on the quantum manifold" scores as easy reading. Short words and a short sentence are all the formula looks at.
- !
They punish useful long words. "Photosynthesis" costs five syllables whether or not your readers learned it in year seven.
- !
Choppy writing scores well. Cut every sentence to eight words and the grade level drops while the prose becomes exhausting to read.
- !
Short samples are noisy. Under 200 words, one unusual sentence can move the grade level by several years. Treat brief results as a rough hint.
Origins of the Formulas
The numbers you see here are not arbitrary. Each formula has a published origin, and knowing that helps you trust (or question) the output:
- Flesch (1948) – Rudolf Flesch, A New Readability Yardstick, Journal of Applied Psychology.
- Flesch-Kincaid (1975) – Kincaid et al., developed for the US Navy training manuals.
- Gunning Fog (1952) – Robert Gunning, The Technique of Clear Writing.
- SMOG (1969) – G. Harry McLaughlin, designed for assessing health literacy.
- Coleman-Liau (1975) – Coleman & Liau, introduced character-based scoring.
- ARI (1967) – Senter & Smith, automated readability index for electronic texts.
Common Mistakes When Using Readability Scores
Even experienced writers trip over these. Avoid them and your scores will actually mean something:
- ✕
Optimizing for the lowest score. A grade 3 text about quantum physics is nonsense. The score must match the complexity of the idea, not the other way around.
- ✕
Rewriting for the formula, not the reader. "Use" is shorter than "utilize" — but if your readers expect "utilize," changing it makes the text feel clumsy.
- ✕
Ignoring sentence-level flags. The overall score can be fine while one 50-word sentence hides in the middle. Always check the flagged sentences.
- ✕
Testing only one formula. If Flesch says grade 8 and SMOG says grade 14, something is off. The cluster of all six is more trustworthy than any single number.
- ✕
Taking the grade as final. A number is a clue, not a verdict. Use it to find the rough patch, then read the text aloud. If it sounds right, it probably is.
What to Remember
A readability score is a smoke alarm, not a critic. It cannot tell you whether your argument holds up, but it will reliably tell you when a sentence has grown too long for anyone to follow. Use the number to find the problem sentences, then trust your own judgment about the fix.
For more writing guidance, check out our guide to common grammatical errors or try the AI writing prompt generator to beat writer's block.
Ready to Check Your Text?
Paste it above. Six scores, instantly, without leaving your browser.
Try It Now — FreeFrequently Asked Questions
Is it really free?
Yes, and there is no daily limit either. The calculation happens on your own device, so there is no server cost for us to pass on to you.
What is a readability score?
A formula that estimates how hard a text is to read, based on measurable features like sentence length and syllables per word. It tells you nothing about whether the writing is good, only how much decoding effort it demands.
What is the Flesch Reading Ease score?
A 0 to 100 scale where higher means easier. Roughly: 90 and above is very easy, 60 to 70 is plain English, and below 30 is dense academic prose.
What is the Flesch-Kincaid Grade Level?
The same underlying measurements expressed as a US school grade instead of a 0 to 100 score. A result of 8.0 means an average eighth grader should be able to follow the text.
Which formula should I trust?
Look at all six rather than one. They weigh sentence length and word complexity differently, so when they cluster around the same grade level you can be confident. When they diverge sharply, something unusual is going on in the text. See the "Which Formula Should You Use?" section below for a detailed breakdown.
How accurate are the scores?
The formulas themselves are applied exactly as published. Syllable counting uses an algorithm that is accurate to within a few percent on ordinary English prose, though it can misjudge unusual names and loanwords. The formulas have a wider margin of error than that, so this rarely changes the picture.
Is my text private?
Completely. Nothing is uploaded. You can disconnect from the internet after the page loads and the tool will still work.
Does it work for non-English texts?
Not reliably. Every one of these formulas was calibrated on English, and syllable counting assumes English spelling patterns. You will get numbers for other languages, but they will not mean much.
How long should my text be?
At least 200 words. Below that, a single long sentence can swing the grade level by several years.
Why does my text score higher than I expected?
These formulas only measure sentence length and word length. Short sentences full of jargon will score as easy even when the ideas are difficult. Use the sentence-level flags alongside the scores rather than relying on the number alone.
Other Useful Tools
AI Report Card Comment Generator
Polished comments from quick notes.
AI Rubric Generator
Analytic, holistic, or single-point rubrics.
AI Response Generator
Professional emails from messy notes.