Keyword density is (keyword occurrences divided by total words) times 100, and there is no officially recommended figure — Google has never published an ideal density and ranks pages on meaning and relevance rather than a keyword count. As a rule of thumb, many SEO professionals treat roughly 0.5 to 3 percent for a primary term as comfortable and natural, and Yoast, for instance, suggests around 0.5 to 3 percent; above about 3 to 4 percent a single term starts to read as repetitive and can look like keyword stuffing. But these are guidelines for spotting accidental over- or under-use, not a target to optimise toward. The better practice is to place your keyword where it naturally fits — the title, the first paragraph, a heading or two — and then write for readers. AnyTool’s tool shows your density and flags terms above ~4% so you can catch over-use, without pretending there is a perfect number.

Keyword stuffing means filling a page with a keyword — or lists of words, phone numbers or place names — in an unnatural, repetitive way to try to manipulate search rankings. Google’s web spam policies explicitly list it as a violation: its automated systems and, where needed, human reviewers can respond, so a stuffed page may rank lower or, in severe cases (especially invisible techniques like hidden text or the same phrase repeated over and over), be removed from results entirely. It also does not even work anymore — you do not need to use a keyword a set number of times to rank for it, and modern engines treat repetitive or irrelevant keywords as spam. The fix is to write useful, information-rich content that uses keywords in context. This tool flags single words above ~4% density precisely so you can catch accidental stuffing before it becomes a problem.

TF-IDF stands for term frequency times inverse document frequency. Term frequency is simply how often a word appears in your page; inverse document frequency measures how rare that word is across a large set of documents, so words that appear on almost every page (like “best”, “guide” or “and”) get down-weighted, while distinctive topic words get boosted. Multiply the two and you get a score that highlights the special vocabulary of your content rather than filler. The catch is that a true IDF needs a corpus — many documents — to know what is rare, and a single pasted document cannot provide that. So this tool approximates IDF using a built-in reference of the most common English words: terms near the top of that list score low, and words absent from it score high. That makes the TF-IDF column here an honest approximation that surfaces distinctive terms, not a true corpus TF-IDF, which is why it is best used to compare the relative prominence of terms in your own text.

The Keyword Density & TF-IDF tool is 100% client-side: the text you paste is split into words and phrases and counted with ordinary JavaScript on your device — nothing is fetched, uploaded, logged or stored on a server, and no CDN is called for processing. That makes it safe for analysing unpublished articles, client copy and confidential content without exposing them. It analyses as you type, offers 1-, 2- and 3-word phrase tables, a stop-word toggle, a target-keyword density check and a copyable report; it works offline once cached as a PWA, supports dark mode and a mobile layout with a sticky Sample / Copy bar. The logic is a small pure engine (keywordDensityEngine.ts) that reuses the shared textStatsEngine tokenizer with no new dependency — everything runs locally in your browser.