BUY WEBSITES
BUY DOMAINS
LOGIN
SIGNUP
N-Gram Phrase Frequency Tool
An n-gram phrase frequency analysis counts the contiguous multi-word phrases that repeat across a block of text, so the recurring 2, 3, and 4-word combinations a page actually leans on become visible instead of guessed at. Where keyword density looks at single words, phrase frequency surfaces the bigrams and trigrams that signal topic and intent, which is what search engines parse and what writers unconsciously overuse.
This tool tokenizes your pasted text, builds every contiguous n-gram inside each sentence, counts them, and ranks the most common phrases with a minimum-frequency filter and a stopword toggle. It is completely free and browser-based, so the text you analyze is never uploaded, logged, or sent to any server.
| # | Phrase | Count | % of phrases |
|---|
🔒 Private: everything runs in your browser. Nothing you paste is uploaded.
How to use the n-gram phrase frequency tool
Paste the copy you want to audit, choose a phrase length, set the minimum count, and read the ranked table. The results recompute on every change, entirely in your browser.
Choose the phrase length that matches what you are checking
Bigrams (two words) expose brand names, product terms, and the repeated word pairs that define a page’s core topic, which makes them the fastest way to confirm a piece is on-subject. Trigrams and four-grams catch longer-tail patterns and the stock phrasings a writer falls back on, such as a recurring call-to-action or a boilerplate sentence opener. Start at two words to read the topical signal, then step up to three or four to find longer repeated strings worth rewriting for variety. Each length is counted independently, so switching the selector recomputes the whole table.
Use the minimum-frequency filter to separate signal from noise
Most phrases in any text occur exactly once, and a list of one-off n-grams tells you very little. Raising the minimum count to two or three collapses that noise and leaves only the phrases a page genuinely repeats, which are the ones worth acting on. A bigram that appears seven times in a short article is a deliberate or accidental focus phrase; the filter is what makes it stand out. Lower the threshold when auditing short copy and raise it for long pages, then read the count column rather than the percentage to judge real repetition.
Toggle stopwords to keep the table meaningful
Function words like the, of, and to bind sentences together but carry no topical meaning, so phrases built entirely from them, such as of the or in a, dominate a raw frequency count and bury the phrases you care about. The stopword toggle removes only the n-grams that are wholly stopwords, while keeping mixed phrases like the backlink profile that still contain a content word. Leave it on for a clean topical read, and switch it off only when you specifically want to inspect grammatical patterns or sentence scaffolding rather than subject matter.
N-gram phrase frequency frequently asked questions
Q1What is an n-gram in SEO?
An n-gram is a contiguous sequence of n words taken from a text. A 2-gram, or bigram, is a two-word phrase, a 3-gram is three words, and so on. In SEO, counting n-grams reveals which multi-word phrases a page repeats, which helps confirm topical focus, spot accidental over-repetition, and understand the phrasing a competitor leans on across their content.
Q2How is phrase frequency different from keyword density?
Keyword density measures single words as a percentage of the total, while phrase frequency counts multi-word phrases and lets you filter by how often each one repeats. A single word can look balanced while a two or three-word phrase is badly overused. Phrase frequency with a minimum-count filter surfaces that pattern, making it the more useful view for editing and topical analysis.
Q3Why does the tool not count phrases across sentences?
A phrase that spans a sentence break, joining the last word of one sentence to the first word of the next, is not a real phrase a reader perceives. The tool splits text on sentence punctuation before building n-grams, so every counted phrase sits inside a single sentence. This keeps the counts meaningful and matches how language models and search engines treat phrase boundaries.
Q4What does the stopword toggle actually remove?
It removes only n-grams composed entirely of common function words, such as of the, in a, or to be. Any phrase containing at least one content word is kept, so the backlink or aged domain still appear. This strips out grammatical filler that would otherwise top the chart while preserving every phrase that carries topical meaning, giving a cleaner read of what the text is about.
Q5Is my pasted text uploaded anywhere?
No. The entire analysis, including tokenizing, building n-grams, counting, and sorting, runs in your browser using JavaScript. Nothing you paste is sent to a server, logged, or stored, and the tool keeps working offline once the page has loaded. That makes it safe for analyzing unpublished drafts, client documents, or any confidential text.
