N-Gram Generator

An n-gram is a sequence of n consecutive words taken from a text. A unigram is one word, a bigram is two ("search engine"), and a trigram is three ("best running shoes"). Linguists use n-grams to model how language is used, search engines use them to understand phrases, and writers can use them to see which ideas a text keeps returning to.

This tool slides a window across your text one word at a time, records every two- and three-word sequence, and counts how often each one appears. Matching ignores capital letters, so "Running Shoes" and "running shoes" are counted together.

Search engines no longer reward repeating an exact keyword, but the phrases a page uses still signal what it is about. N-gram analysis helps you:

• Check topical focus. The top bigrams on a page should describe its main subject. If they do not, the page may be drifting.
• Avoid over-optimisation. A single phrase repeated far more than anything else reads badly and can look like keyword stuffing.
• Research competitors. Paste the text of a page that ranks well to see which supporting phrases it covers that yours does not.
• Find long-tail ideas. Trigrams often reveal specific questions and modifiers worth their own section or article.

For single-word percentages, use the Keyword Density Checker.

Every writer has habits: a pet phrase, a favourite transition, an opening that keeps coming back. Those patterns are hard to see when you read your own work, but they jump out of an n-gram table. Look for phrases such as "it is important to", "in order to" or "at the end of the day" appearing several times, then vary or cut them.

To find repeated single words and accidental doubles like "the the", use the Duplicate Word Finder.

Common function words ("of the", "in a") naturally rank highly in any English text; skip past them to the phrases that carry meaning. Results are capped at the ten most frequent phrases in each table. Punctuation is ignored, so phrases can run across sentence boundaries. That is normal for n-gram counting and rarely affects the top results on a real page.