SEO Text Analyzer | Keyword Density Checker

Term	Count	Density
Paste some text to see its top terms.

Paste a draft to see which terms actually dominate it. The analyser counts single words, two-word and three-word phrases, skipping stop words like "the" and "and" so the list reflects the topic rather than English grammar.

Density is a diagnostic, not a target. Modern search ranking does not reward a magic percentage, but a page about "cold brew coffee" that never repeats the phrase is a signal something is off, and one that repeats it forty times reads like spam to humans, which is the audience that matters.

How to use it

Paste the page copy you want to analyse.
Switch between one-, two- and three-word phrases.
Compare the top terms against the keyword you are actually targeting.

Questions

What keyword density should I aim for?

There is no correct number, and no search engine publishes one. Roughly 0.5-2.5% for a primary term reads naturally; well above that reads like keyword stuffing to a person, which is the real risk.

Why are common words missing from the list?

Single-word results skip a stop-word list (the, and, with, that and similar) because they top every text and tell you nothing. Two- and three-word phrases keep them, since 'how to brew' is a meaningful phrase.

Does this check my live page?

No, it analyses text you paste. That keeps it instant and private; to audit a published URL you need a crawler that fetches the page.

Do meta keywords still do anything?

No. Google confirmed publicly in 2009 that it ignores the meta keywords tag for ranking, and nothing has changed since. The one thing it still does is tell a competitor exactly which terms you are targeting, which is a strange gift to leave in your markup.

Where should the target phrase actually appear?

The title, the H1, somewhere in the first hundred words, and one subheading, all in sentences that would exist anyway. After that, stop. Repetition beyond what the writing needs adds nothing and costs you readers, and the analyser is here to show you when you have crossed that line without noticing.

Is there a word count that ranks better?

No, and the studies people cite for one are measuring correlation. Longer pages tend to rank better because covering a subject properly usually takes more words, not because length itself is rewarded. Padding a 400-word answer to 2000 words makes it worse in the only way that matters, which is whether the reader gets what they came for.

What is TF-IDF, and is that what this measures?

It is a way of weighting a term by how often it appears here against how rare it is everywhere else, so common words score low without needing a stop-word list. This page measures plain density and frequency instead, because TF-IDF needs a corpus to compare against and there is not one inside your browser tab.

Can I analyse a page that is already published?

Not directly, since the analyser reads text you paste rather than fetching a URL, which is what keeps it instant and private. Copy the visible copy from the page, or view the source and take the body text. Auditing a live URL properly means a crawler that can see the rendered HTML, the headers and the redirects.

The lengths that get truncated

Element	Practical limit	What happens past it
Title tag	About 60 characters	Cut in the result, usually mid-word
Meta description	About 155 characters	Cut, or replaced entirely by Google
H1	One per page	Nothing breaks, but the page stops having a clear subject
URL slug	3 to 5 words	Still works, reads worse when shared
Image alt text	About 125 characters	Screen readers keep going, listeners stop
Open Graph title	About 60 characters	Truncated in the link preview

Titles are actually measured in pixels rather than characters, roughly 580 of them on desktop, so a title full of capitals and wide letters truncates earlier than the character count suggests. Sixty is a safe approximation, not a rule.

Meta descriptions are worth writing anyway even though Google rewrites most of them, because it rewrites them from the page, and a page with a clear summary gives it something better to choose.

Your data stays on your device

Everything above runs inside your browser as WebAssembly compiled from Rust. Nothing you type is uploaded, logged or stored on a server. You can load this page once, go offline, and it still works.

This page makes no requests at all, to anywhere. That is not a promise in the copy: it is a Content-Security-Policy header your browser enforces, and connect-src on it is none. Open the network tab and watch nothing happen.