Summary

This free keyword clustering tool groups a list of up to 300 keywords by shared wording, entirely in your browser. It drops filler and intent words, reduces each keyword to word stems, weights rare words above common ones, and puts keywords with enough overlap in the same group. Each keyword also gets an intent guess (informational, commercial or transactional), and groups that mix intents are flagged. It does not read Google results, so it is a first sort to check, not a final decision.

A Free Keyword Clustering Tool That Shows Its Rules

Paste a keyword list and get groups you can map to pages, with an intent guess for each keyword. The rules are simple enough to read and run entirely in your browser.

Cluster my keywords
Index cards sorted into four paper-clipped groups on a wooden desk

Group your keywords in seconds

Paste up to 300 keywords, one per line or separated by commas. Groups update as you type. The first keyword of each group is the suggested page target.

This tool groups keywords by shared wording only. It cannot see Google results, so two keywords that look different but return the same pages will land in different groups. Treat the output as a first sort, then check the borderline groups against the live search results.

How it works

What happens to your list

Clean and stem each keyword

Each keyword is lowercased, stripped of filler words ("the", "for", "with") and of intent words such as "best", "how" or "buy". Endings are cut so "writing", "writer" and "write" count as the same word.

Weighted overlap, not raw overlap

Two keywords are compared by the stems they share. A word found in most of your list, like "keyword" or "tool", counts for less than a rare one, so "keyword research tool" does not merge with "keyword clustering tool" just because of the two generic words.

A guess at intent

The words stripped in step one also feed an intent label. When one group holds informational and commercial keywords together, the tool flags it, because those often need separate pages.

Know the limits

Text similarity is a first sort, not a verdict

There are two common ways to cluster keywords. The first compares the words themselves, which is what this page does. The second fetches the search results for each keyword and groups the ones that return the same pages. One vendor comparison published in April 2026 describes the usual rule as three or more shared pages in the top ten. That second method sees what Google sees, so it catches pairs the first one cannot, and it costs search data to run. Use this tool to get a fast, readable draft of your groups, then check the groups you doubt against the live results before you commit one page to each.

  • Good at: plurals, word order, filler words, prefixes like "best" or "free"
  • Blind to: synonyms such as "tool" and "software", and shared intent hidden behind different words
  • Best used on lists of 20 to 300 keywords, as a first pass before a manual review
Hands sorting blank cards into columns on a light wood table
After clustering

From groups to a publishing plan

  1. 1

    Pick one target keyword per group

    The first keyword of each group is the shortest, so it is usually the broadest. Keep it if it matches what you want the page to rank for, or swap in the keyword with the most search volume.

  2. 2

    Split the mixed-intent groups

    If a group is flagged as mixed intent, read its keywords. A "how to" keyword and a "best" keyword in the same group often want an explainer and a roundup, not one page.

  3. 3

    Map each group to a page

    Export the groups as CSV, add a column for the URL you will use, and mark the groups with no matching page. Those are your next articles.

Common questions

Is this keyword clustering tool really free?
Yes. There is no sign-up, no email gate and no keyword cap beyond 300 per run. The grouping runs as JavaScript in your browser, so there is nothing on our side to meter.
Does it use Google search results to group keywords?
No. It groups by wording only: shared stems, weighted by how rare each word is in your list. Paid clusterers such as Keyword Insights and KeyClusters fetch live results and group keywords that return the same pages. That is a different method, and a more reliable one for deciding which keywords share a page.
How accurate are the groups compared with a SERP-based tool?
Accurate for obvious variants (plurals, word order, filler words, "best" or "free" prefixes), weaker on synonyms and on keywords that look different but share an intent. We have no benchmark against SERP tools, so treat the groups as a first sort and check borderline ones in the live results.
What do the Informational, Commercial and Transactional labels mean?
They come from a short modifier list. "How", "what", "guide" and "template" read as informational. "Best", "review", "vs" and "alternatives" read as commercial. "Buy", "price" and "trial" read as transactional. A keyword with none of those words is Untagged. It is a hint, not a search intent analysis.
Why did two similar keywords end up in different groups?
Usually because they use different words for the same idea, such as "tool" and "software", or "writer" and "generator". The tool does not know synonyms. Move the strictness to Loose, or merge the two groups by hand.
What does the strictness setting change?
It sets how much two keywords must overlap to share a group. Strict needs a similarity of 0.6, Balanced 0.45, Loose 0.3. Lower values give fewer, larger groups and more chance of unrelated keywords sharing one.
Does it work in languages other than English?
It works on any language that separates words with spaces, but the stemming rules are English only, so plural and verb endings in other languages will not be merged. Languages written without spaces, such as Japanese, are not split into words.
Is my keyword list stored or sent anywhere?
No. The list stays in your browser tab. The only thing sent is an anonymous tool-run signal that carries no keywords and no personal data.

Want more honest tests of AI writing tools?

aiContentzy reviews and compares AI copywriting tools on real briefs, no fluff, no sponsored puffery.

Browse aiContentzy
aiContentzy
Open the generator