Keyword Extractor
FREE TOOL · NO SIGN-UP
A keyword extractor scans a block of text and lists the words and short phrases that appear most often, after filtering out filler words like “the” and “and”. Paste an article, a competitor page or a transcript into the tool above, choose single words, 2-word or 3-word phrases, and you get a ranked frequency table plus a plain keyword list you can copy. It shows what a text is really about, which makes it useful for content audits, briefs and gap checks.
What Keyword Extraction Does
Keyword extraction is a basic form of text analysis. The tool breaks your text into tokens, counts how often each term repeats, and sorts the results so the core topics rise to the top. A page about running shoes should surface terms like “running shoes”, “cushioning” and “heel drop”. If it surfaces “click here” and “our team” instead, the copy has drifted away from its subject.
That makes extraction a fast sanity check for content strategy. Writers use it to confirm a draft covers the topic they planned. Editors use it to spot repeated phrases that make copy sound stiff. SEOs paste in competitor content to see which terms the ranking pages keep returning to.
What extraction does not do is tell you how people search. It reads only the text you give it. Treat the output as a map of what a document says, then check that map against real keyword research before you change a page.
How the Extractor Finds Terms
The FHSEO Hub extractor runs on term frequency. It lowercases the text, splits it into words, drops anything shorter than your minimum word length, and counts every remaining word or word sequence (n-grams). You control four settings: how many keywords to show (top 10, 25, 50 or 100), the minimum word length (2, 3 or 4 characters), whether stop words are removed, and whether numbers count as keywords.
Results appear as a table with the rank, the keyword, how many words it contains, its frequency and a bar showing its weight against the top term. Below that sits a one-per-line list you can copy or download as a text file.
Single words, bigrams and trigrams
Single words show the broad vocabulary of a page. They are good for spotting the main entity but poor at showing context, because “content” alone could mean content marketing, content gaps or content management.
Bigrams (2-word phrases) and trigrams (3-word phrases) fix that. “Content strategy” and “topic clusters” tell you far more than the single words inside them. They also tend to look like the queries people type into search engines.
The tool only lists a 2-word or 3-word phrase if it appears at least twice, so one-off word pairs do not clutter the table. Pick “All” to see words and phrases ranked together. In that view, longer phrases get extra weight, so a trigram that repeats a few times can rank above a common single word.
Why stop words are removed
Stop words are the noise words of a language: “the”, “is”, “of”, “to”, “and” and similar. In any normal paragraph they are the most frequent terms by a wide margin, so leaving them in pushes every meaningful keyword down the list.
With “Remove stop words” ticked (the default), those words are filtered out before counting. Untick it when stop words carry meaning for your analysis, for example when you want to see exact phrases such as “how to” or “for beginners” that often appear in question-style queries.
How to Extract Keywords in Three Steps
Paste your text. Copy the body copy of an article, product page, video transcript or competitor page into the content box. Leave out menus, footers and cookie notices, since repeated navigation labels will distort the counts. The tool does not fetch URLs, so copy the text from the page yourself.
Set the filters. Choose how many keywords you want, set the minimum word length to 3 or 4 to skip short fragments, and decide on stop words and numbers. Turn numbers on if model names, years or sizes matter (for example “iphone 15” or “2026 guide”).
Switch phrase length and read the table. Start with single words to see the main subject, then switch to 2-word and 3-word phrases to find the actual topics. When you have a list you like, press Copy or Download to save it as keywords.txt for a spreadsheet or a content brief.
The results update as you type or change a setting, so you can tweak filters and compare views quickly.
Turning Extracted Terms Into Content Strategy
A frequency list is raw material. It becomes useful when you compare it with something else: a competitor page, your target query or your own older content. Two workflows cover most needs.
Spotting content gaps against competitor pages
Take the top three ranking pages for your target query. Paste each one into the extractor, switch to 2-word phrases and save the lists. Then run your own page the same way.
Terms that appear on all three competitor lists but not on yours are candidate content gaps. They often point to subtopics searchers expect, such as “return policy”, “size chart” or “setup time”. Terms that appear only on your page may be the angle that sets you apart, or they may be off-topic padding.
Do not copy the competitor vocabulary wholesale. Use the gaps as questions: does my page answer this? If not, add a section that covers the subtopic properly rather than dropping the phrase in once.
Grouping phrases into topic clusters
Extracted bigrams and trigrams often fall into natural groups. A long guide on email marketing might return phrases about deliverability, list building and subject lines. Each group can become a section of the page or a supporting article in a topic cluster.
Paste the downloaded list into a spreadsheet and tag each phrase with a group label. Phrases that share a head term (“email list”, “email list growth”, “clean email list”) usually belong together. For bigger lists, the Keyword Grouper can sort phrases by shared words for you, and the Duplicate Line Remover cleans up overlaps after you merge lists from several pages.
Frequency Versus Search Volume
Frequency tells you how often a term appears in one document. Search volume tells you how often people type a query into Google. The two are unrelated. A page can repeat “innovative solutions” twenty times while nobody searches for it, and a phrase that appears once might be a query with steady demand.
Use extraction to understand a document and keyword research tools to understand demand. A practical order is: extract terms from top-ranking pages, shortlist the phrases that look like real queries, then check those phrases in a keyword research tool or in Google Search Console. Keep the ones with demand and clear search intent.
Mistakes When Reading the Output
Treating the top word as the target keyword. The most frequent single word is often a generic term like “content” or “product”. Look at phrases for the real topic.
Pasting the whole page source. Navigation, sidebars and footer links repeat on every page and inflate terms like “home”, “contact” or “read more”. Paste the main body only.
Ignoring word variants. The extractor counts exact word forms. “Shoe” and “shoes” are counted separately because it does not perform lemmatization. Add the counts together in your head or in a spreadsheet when you judge a topic.
Reading counts without length. A term repeated eight times in 300 words is heavy; the same count in 3,000 words is light. Use the Keyword Density Checker when you need percentages.
Chasing a frequency number. Search engines do not reward hitting a set count. Repeating a phrase to climb the table makes copy worse to read and can look like stuffing.
Where Frequency-Based Extraction Is Wrong
Counting words is simple and transparent, but it misses meaning. A short page that mentions its main topic only twice can still be about that topic, while a frequent term might be a side tangent.
Statistical methods such as TF-IDF compare a document against a larger collection and reward terms that are rare elsewhere. Natural language processing models, including AI tools like ChatGPT, can pull out concepts that never appear as exact repeated strings. This extractor does neither. It shows plain counts, which makes it predictable but blind to synonyms, sarcasm and context.
Use it for what it is good at: a quick, honest picture of the words a text actually repeats.
Private Text Analysis in the Browser
The extractor runs entirely in your browser. Your text is processed on your own device and is not uploaded to FHSEO Hub or stored on a server. That makes it safe to paste unpublished drafts, client copy or internal documents. Clearing the page or pressing Clear removes the text from the tool.
Pairing Extraction With Real Search Data
Extraction tells you what a page says; search data tells you what people want. Run both. Pull phrases from top-ranking pages here, check demand in your keyword tool, then combine seed terms with the Keyword Combiner. You can find more free tools for every step in the FHSEO Hub free SEO tools collection.