Readability & Content Analyzer

Paste a URL to get its word count, reading level, sentence length, text-to-HTML ratio and most-used words — the content numbers that decide whether a page is worth ranking. Free, instant, no signup.

What this measures & what good looks like

It loads the page's rendered HTML, strips scripts, styles, SVG and other non-text nodes, and picks the content region: <main> if present, otherwise <article>, otherwise the body with navigation, header, footer and sidebars removed. From that text it counts words, sentences (two or more words ending in terminal punctuation or a block boundary) and <p> paragraphs, measures average sentence and word length, flags sentences over 25 words, and estimates reading time at 238 words per minute.

Flesch Reading Ease and Flesch–Kincaid Grade use the standard formulas with a syllable heuristic (vowel groups, with silent-e handling), and are only reported when at least half the words are Latin-script and there are 30+ words in 3+ sentences, since the formulas are calibrated on English. Text-to-HTML ratio compares the visible text of the whole page to the raw HTML size. The top-words list excludes stopwords and anything under three characters, so what remains is the vocabulary a search engine would actually associate with the page.

What good looks like

  • Flesch 60–70 (grade 8–9) for a general audience; 50–60 is fine for technical documentation; below 30 is academic or legal prose and will lose most web readers.
  • Fewer than 15% of sentences over 25 words. Long sentences are not wrong, but each one is a place a reader can lose the thread — split at the conjunction.
  • 300+ words for a page you expect to rank for an informational query. Pages under 100 words rarely rank on their own; that is fine for a contact page, not for an article.
  • Text-to-HTML above 10%. Below 5% usually means markup, inline scripts or a hydration payload are dwarfing the content — check the page weight too.

Frequently asked questions

Is readability a Google ranking factor?

Not directly — Google has said it does not compute a readability score and does not reward simple text for its own sake. What it does measure is whether people who land on a page stay, scroll and find what they came for, and prose that is hard to parse loses them. Readability is also what makes a passage quotable in featured snippets and AI answers, which are increasingly where clicks come from. Write for the reader; the metrics are a way of checking you have.

Why does it say the reading level was not measured?

Flesch Reading Ease and Flesch–Kincaid were built on English syllable and sentence patterns, so the tool only applies them when at least half the words on the page use Latin script and there are 30 or more words across three or more sentences. Pages in Hindi, Arabic, Chinese, Japanese and other non-Latin scripts get every other statistic — word count, sentence length, reading time, top words — but no Flesch score, because the number would be meaningless.

How is reading time calculated?

Word count divided by 238 words per minute, the average adult silent-reading speed for non-fiction found in a 2019 meta-analysis of 190 reading studies. Medium uses 265 and some tools use 200; none of them are wrong, they just assume different readers. Technical material with code samples or tables tends to take longer than the estimate.

What is text-to-HTML ratio and does it matter?

The bytes of visible text on the page divided by the bytes of raw HTML, as a percentage. It is not a ranking factor and there is no magic threshold, but it is a quick diagnostic: a page at 3% is either very light on words or very heavy on markup, inline scripts and hydration data, and both are worth a look. Content pages usually land between 10% and 30%; app-like pages sit lower without it being a problem.

Why is the word count lower than what my CMS or editor reports?

The tool analyses the main content region — the <main> or <article> element when the page has one, otherwise the body with navigation, header, footer and sidebars stripped out — so menu labels, cookie banners and footer links are not counted. Editors count everything you typed, including headings and captions, and some count hyphenated compounds as several words. A difference of 5–15% is normal; a much larger gap usually means the page's main content is not in a semantic container the tool can find.

Related tools

Meta Tag AnalyzerHeading Structure CheckerImage Alt CheckerKeyword Density

Command Palette

Search for a command to run...