What Keyword Density Checker does
The Keyword Density Checker fetches one public HTML page, extracts visible text, lowercases and tokenizes words, removes a small English stop-word list, and counts phrase windows. It returns up to twenty-five results for single words, two-word phrases, and three-word phrases.
Density is calculated against the usable phrase windows for each phrase size, so the percentages help compare repetition within this extracted sample. They are not SEO targets. Natural language, page purpose, semantic coverage, and usefulness matter more than reaching a fixed percentage, and JavaScript-rendered or inaccessible text may not appear in the report.
How to use Keyword Density Checker
- Enter the full public URL of the page you want to inspect.
- Select Analyze density and wait for the bounded HTML fetch and text analysis.
- Compare counts and density percentages across the one-, two-, and three-word lists.
- Review repeated phrases in their actual page context before editing copy.
Important details
Phrase density uses usable windows
For each phrase size, the denominator is the number of phrase windows that remain after stop-word filtering. A two-word density is therefore comparable within its own list, not directly interchangeable with a single-word percentage.
Only bounded visible HTML text is counted
Scripts, styles, markup, and a small English stop-word set are excluded. The fetch does not execute page JavaScript, so client-rendered copy may be absent and very large input is capped.
Frequency is descriptive, not prescriptive
A high count can reveal accidental repetition, navigation labels, or a genuinely central topic. The tool does not label a percentage good or bad and does not predict a search engine response.
Practical uses
- Spot unusually repeated terms before publishing a landing page.
- Compare the language emphasis of two versions of the same page.
- Identify common multi-word phrases for an editorial review, not as ranking guarantees.
Privacy and limitations
The URL is sent to PagesTools, whose server fetches the public page and analyzes its text. The target site can log the server request; the report is returned with no-store cache headers.
The analysis covers one bounded, server-fetched HTML response, excludes JavaScript-rendered content, uses a small English stop-word list, and provides no search volume, intent, competition, stemming, or ranking data.