Keyword Extractor
Paste text and extract the most repeated meaningful words and phrases with counts and percentages. Local text-frequency analysis only.
Your input
Everything you type is processed in this browser tab. Nothing is uploaded, logged or stored.
Results
Fill in the fields and choose Run tool to see results here.
About this tool
Extraction answers a simple question: what is this text actually about? By counting the words and phrases that repeat, you get an honest picture of a page's emphasis — your own draft, a transcript, a competitor's copy you are studying, or a description you need to summarise. Everything is computed in the browser from the text you paste, and nothing is stored.
How it works
The text is lowercased and split into word tokens. For single-word mode, common stop words such as the, and and with are removed along with very short tokens, then each remaining word is counted. For two- and three-word modes, the tool slides a window across the text and counts each n-gram, discarding phrases that begin or end with a stop word so that fragments like of the do not dominate. Results are ranked by count and shown with their share of the analysed text.
- 1Paste the text you want to understand.
- 2Choose whether to extract single words, two-word or three-word phrases.
- 3Set how many results to return.
- 4Generate, then copy the table or download it as CSV.
Example
Input
An article that repeats 'keyword research' four times
Output
keyword research — 4 — 6.15%
When to use it
- Checking whether your draft emphasises the topic you intended.
- Summarising a long transcript into its recurring themes.
- Studying how a well-performing article distributes its language.
- Pulling candidate phrases out of customer feedback or reviews.
Tips
- Two-word mode is usually the most useful starting point for real topics.
- Compare your draft against a strong reference article and look at the differences, not the absolutes.
- Never edit a page just to raise a percentage — write for the reader first.
Limitations
Frequency is not importance and it is definitely not search volume. A phrase can appear ten times and still be irrelevant, and a crucial phrase might appear once. The analysis is also language-aware only for English stop words, so results in other languages will include more filler terms.
Frequently asked questions
- Does a high percentage mean better SEO?
- No. There is no target density that improves rankings, and stuffing a phrase to hit a number reliably makes copy worse. Use the numbers descriptively, not as a goal.
- Why are some obvious phrases missing?
- Phrases that start or end with a stop word are filtered out to keep results meaningful, and single-word mode ignores words of three characters or fewer.
- Is there a length limit?
- There is no hard limit, but very long documents take longer because everything is processed in your browser. A few thousand words is comfortable.