Repeated Phrase Finder

How to use

1. 1Paste or type your text into the input area on the left, or click Load sample to try the tool on a worked example with obvious repetition.
2. 2Set the minimum and maximum phrase length (two to ten words) to choose how long the phrases you are hunting for can be.
3. 3Set Minimum repeats to control how often a phrase must appear before it is listed, and turn Case sensitive on if capitalization should matter.
4. 4Keep Collapse sub-phrases on to hide shorter fragments that sit inside a longer repeated phrase, or turn it off to see every repeated phrase of every length.
5. 5Sort by repeats, phrase length, or first appearance, use Find a phrase to filter, and click Copy CSV or Copy plain text to export the list.

About this tool

The Repeated Phrase Finder scans a block of text and surfaces the multi-word phrases that show up more than once, ranked by how often they repeat. It answers a specific question that a plain word counter cannot: not which single words you use most, but which exact runs of words (two, three, four, or more in a row) you keep reusing. Paste an essay, article, blog post, transcript, cover letter, thesis chapter, product description, or any prose, and the tool tokenizes the text into words, then slides a window of every length in your chosen range across the text and counts each phrase. Tokenization is Unicode-aware (accented Latin, Cyrillic, Greek, and other scripts tokenize cleanly) and keeps apostrophes inside words so contractions like don't and it's stay intact, and curly apostrophes are normalized so the same phrase does not split into two. Phrases are never counted across sentence boundaries: the text is divided at sentence terminators and structural punctuation first, so the trailing words of one sentence and the opening words of the next are not glued into a phantom phrase. You control the analysis with a few options. A minimum and maximum phrase length (from two up to ten words) sets the window range. A minimum repeats value hides phrases that do not appear often enough. A case-sensitive toggle decides whether The End and the end count as the same phrase. The most important quality control is collapse sub-phrases: when a long phrase repeats, every shorter phrase inside it technically repeats too, which would otherwise flood the list with fragments of the same passage, so this option hides a shorter phrase whenever it is fully contained inside a longer phrase that repeats at least as often, leaving the longest meaningful unit. A find-a-phrase search filters the results by substring. Each result row shows the phrase exactly as it first appears, how many times it occurs, how many words it spans, a relative bar, and a copy button, and you can sort by number of repeats, phrase length, or first appearance, then export the whole list as CSV (Phrase, Words, Count) or as aligned plain text. It is built for writers and editors trimming accidental repetition, students checking that a paper does not lean on the same stock phrases, content teams catching duplicated boilerplate across a document, speakers and podcasters reviewing transcripts for verbal crutches like at the end of the day or you know what I mean, and translators confirming that a recurring term is rendered consistently. Everything runs locally on your device, so the drafts, transcripts, and documents you paste here never leave your browser.

Free to use. Works in your browser. No signup, no login.