Check how hard your writing is
Paste your text, upload a file, or load the example. Six readability formulas, a grade level, and every sentence graded so you can see which ones are doing the damage.
Paste your writing and six readability formulas score it as you type — Flesch Reading Ease, Flesch–Kincaid, Gunning Fog, SMOG, Coleman–Liau and the Automated Readability Index. Then it does the part a score cannot: grades every sentence on its own, lists the hardest first, and names the words that pushed the number up. Free, no sign-up, nothing leaves your browser.
Score, then rank, then edit.
Type, paste, or drop a .docx, .pdf or .txt file onto the editor. Files are read in your browser; nothing is uploaded.
Five reading levels, from early readers to academic. The verdict is measured against the one you choose rather than a pass mark we invented.
The Sentences tab ranks your sentences hardest first. Click one and the cursor lands on it. That is usually the whole job.
All six read the same two things — how long your sentences are and how long your words are. None of them reads the words. Knowing which lever each one pulls is the difference between using a score and obeying it.
| Formula | Published | Measures length by | Answers in |
|---|---|---|---|
| Flesch Reading Ease | Rudolf Flesch, 1948 | Syllables per word, words per sentence | 0–100, higher is easier |
| Flesch–Kincaid Grade | For the US Navy, 1975 | The same two inputs, rescaled | A US school grade |
| Gunning Fog Index | Robert Gunning, 1952 | The share of words with three syllables or more | A US school grade |
| SMOG Index | G. Harry McLaughlin, 1969 | The square root of that same count | A US school grade |
| Coleman–Liau | Coleman and Liau, 1975 | Letters per word, not syllables | A US school grade |
| Automated Readability Index | For the US military, 1960s | Characters per word, words per sentence | A US school grade |
Coleman–Liau is the one worth understanding. It counts letters rather than syllables, so it is kinder to a long word that is easy to say and harsher on a short one that is not. That is why it is often the formula that disagrees most with the other four, and why a single score from a single formula is a weaker claim than it looks.
A real run, recorded on 26 September 2026. The text on the left is the kind of paragraph that gets circulated at work; the text on the right is the same information after acting on what the tool said. Every figure below came out of the engine on this page.
“Notwithstanding the aforementioned considerations, it is imperative that all personnel remain cognisant of the fact that the implementation of the revised operational framework, which was approved by the committee following an extensive period of consultation with stakeholders across every affected department, will necessitate a substantial reconfiguration of existing reporting lines. Managers are advised. The transition will be completed by March.”
60 words, 3 sentences, 20.0 words a sentence, and 23 of those 60 words carry three syllables or more — 38%.
“The committee has approved the new reporting framework after consulting every affected department. It changes who reports to whom. Managers: please read the summary. We finish the move in March.”
30 words, 4 sentences, 7.5 words a sentence. Nothing was removed except the padding.
The point of the example is which number was useful. “Grade 20” told us nothing we could act on. “One sentence, 50 of your 60 words, graded 30.8” told us exactly what to open. That is the difference between a readability score and a readability checker.
A document score is an average, and averages hide the thing you were looking for.
Readability formulas are driven by average sentence length, so a single 50-word sentence drags a whole page down. Rewrite it and the document score moves by several grades without another word changing. The list finds it in one glance.
Reading your own draft top to bottom is exactly how you miss the problem — you already know what you meant. Sorted by grade, the sentence you should open is the first one on the list.
Every row selects that sentence in the editor. No hunting for the phrase, no scrolling, no second window. Fix it and every score updates as you type.
This matters more than the number, and almost nothing in this category says it.
Pick the audience in the tool and the verdict is measured against that row instead of a number we made up.
| Writing for | Grade aimed at | Roughly Flesch | Where the convention comes from |
|---|---|---|---|
| Children and early readers | 5 | 90+ | Children’s publishing and early-years teaching |
| A general public audience | 8 | 60–70 | The level most newspapers and public-information writing aim at |
| Web and marketing copy | 9 | 60–70 | Plain-English practice: short sentences, common words, nothing to re-read |
| Business and professional | 11 | 50–60 | An informed reader with forty other things to read today |
| Academic and technical | 14+ | under 50 | Specialist readerships, where density is expected and ambiguity still is not |
These are writing conventions, not rules, and nothing on this page enforces them. If a style guide, a course handbook or a client brief gives you a number, that number beats this table — and the Flesch column is approximate because the grade formulas and Reading Ease do not map onto each other cleanly.
Three jobs, one list of sentences.
A supervisor who says “this is hard to follow” is usually reacting to three sentences, not to the whole chapter. The ranked list finds them, and the per-sentence grade tells you whether a sentence is long, dense, or both. Nothing here rewrites for you, which is the point when the work has to be yours.
Plain English is not a ranking factor, but time on page and people finishing the article are downstream of it. Flesch 60 and up, sentences under about twenty words, and the three-syllable list kept honest will do more than any score target. The SEO keyword tool handles the other half.
Policies, notices, instructions, health information, terms. Plain-English guidance in this territory usually names a grade or a Flesch score, and the passive count is often the faster fix — your application was rejected tells nobody who decided or what to do next.
Six pages, six different questions. They share one engine, so the numbers always agree.
How much did I write? Words, characters, reading time, keyword frequency and these same six scores as part of a wider picture. The place to start if you are writing to a word count.
How did I write it? Every sentence listed and measured, with the repeated openings and the rhythm you cannot hear in your own draft.
How long is it on paper? Your text laid out in the real font at the real size, and counted — not multiplied by 250.
What people ask about reading levels and readability scores.
For most writing aimed at a general audience, a Flesch Reading Ease of 60 or above, which is roughly grade 8 to 9. Plain-English guidance for public information usually asks for the same. Higher is easier: above 80 reads like conversation, below 30 reads like a legal contract. What counts as good depends entirely on who is reading, which is why this tool asks you to pick an audience instead of printing a pass mark.
Paste it into the box above. Every score updates as you type, and the Sentences tab lists your sentences graded worst first. Nothing is uploaded and there is no sign-up.
A formula published by Rudolf Flesch in 1948 that turns average sentence length and average syllables per word into a number from 0 to 100, where higher means easier. It is the most widely used readability measure in English and the one Microsoft Word has shown for decades. It measures only length — it has no idea whether the words are the right ones.
They use the same two inputs and express the answer differently. Reading Ease gives a 0–100 score where higher is easier. Flesch–Kincaid Grade Level, developed for the US Navy in 1975, converts the same measurement into a US school grade, so 8.2 means an eighth-grader should manage it. This tool shows both because people are asked for both.
Whichever one the person asking specified. If nobody specified, use Flesch Reading Ease for general writing and look at the average of the grade formulas rather than any single one — they disagree by two or three grades on the same text, which tells you something about how precise they really are. Gunning Fog and SMOG punish long words hardest; Coleman–Liau counts letters rather than syllables, so it treats a long simple word more kindly.
Because they were fitted to different sets of texts and readers, decades apart, and they weight length differently. Coleman–Liau counts characters per word; Gunning Fog counts words of three syllables or more; SMOG uses the square root of that count. On the same paragraph a three-grade spread between them is normal. That spread is the honest error bar on the whole idea, and it is why the tool averages the five grade formulas rather than presenting one as the answer.
No. These formulas measure sentence length and word length, and nothing else. A page of four-word sentences made of nonsense scores beautifully. Writing for specialists is supposed to use specialist words, and cutting a necessary three-syllable term to please a formula makes the writing worse. Treat the number as a smoke alarm: it tells you to look, not what to do.
The Simple Measure of Gobbledygook, published by G. Harry McLaughlin in 1969 and still common in health communication. It was designed to be run on a sample of thirty sentences or more, so below that the tool labels it an estimate rather than quietly printing a number the formula was never fitted to produce.
Because that is the part you can act on. A document score is an average, and averages hide the problem: in the example this page walks through, one 50-word sentence out of three carried the entire score while the other two were already fine. Fixing that one sentence moved the document from grade 20 to grade 10. A single number would never have told you which one to open.
No, and the tool says so where it shows it. It is Flesch–Kincaid applied to one sentence, which is not what the formula was fitted for. It reliably flags something long or dense, which is all it is used for here. The sentence itself is shown to you so you can judge; the number just decides the running order.
By the rule every usable detector uses: a form of to be followed by a past participle, allowing an adverb or two in between. It will flag she is tired, which is not really passive, and it will miss a participle that is not on its list. That is why they are called likely passive constructions and shown to you in context rather than counted silently. The passive is not an error — it is right whenever the doer is unknown or beside the point.
Yes, the -ly ones, and no formula uses them. They are there because they are the cheapest edit in English: ran quickly is sprinted. Words that end in -ly without being adverbs of manner — only, family, likely, friendly — are excluded, so the count means something.
Yes, with no sign-up and nothing held back for a paid plan. Every formula here is arithmetic on counts, so it runs in your browser and costs us nothing to serve. There is no reason to meter it.
Yes — drop a .docx, .pdf or .txt file onto the editor and the text is extracted in your browser. Old .doc files need saving as .docx first, and a scanned PDF with no text layer comes through empty because there is nothing to extract without OCR.
No. Every score is computed in your browser and no part of this tool sends your text anywhere. Your draft is kept in your own browser storage so a refresh does not lose it, and clearing your browser data removes it.
The Word Counter shows these six scores as part of a wider picture — counts, reading time, keyword frequency, platform limits. It answers how much did I write. This page answers which sentences are the problem: every sentence graded and ranked, the passives and adverbs flagged in place, and the three-syllable words listed so the score becomes an edit list. Both use the same engine, so the scores always agree.
Measuring how hard it reads is one job. These are the others.
Free and unmetered, because it runs in your browser. Plan allowances apply only to the AI tools, and they are on the pricing page.
Upgrade
Drop your file in