Skip to content
ZeroServer.tools

Word Frequency Counter

Count how often each word appears and rank them by frequency.

Total Words: 0Unique Words: 0
Results

No words found. Enter some text above.

Analyzing Word Frequency & Density

Word frequency counters are essential for SEO keyword density checks, readability analysis, writing-style audits, and natural language processing tasks. Toggle case-insensitive mode to fold "The" and "the" into the same bucket. The table is limited to the top 100 words for readability; use the TSV export to save the full list to a spreadsheet. All analysis runs entirely in your browser — no data is sent anywhere.

Related tools: Word Counter · Character Map Explorer

Built and maintained by Meet Shah · Last updated

What this tool is used for

  • Finding which terms a document actually leans on.
  • Checking a draft for an unintentionally repeated word.
  • Comparing two documents' vocabulary to see which one is narrower.
  • Producing a ranked term list from survey responses.
  • Spotting filler words that appear far more than you expected.

Frequently Asked Questions

Why are common words excluded?
Because without stopword removal the top of every list is the, of, and, to — true of all English text and therefore uninformative. Filtering them is what makes the remaining counts describe THIS document rather than the language.
Should words be stemmed or lemmatised?
It depends what you are counting. Stemming crudely chops suffixes (running → run), lemmatisation uses a dictionary (better → good). Without either, run, runs and running count as three unrelated words.
Is keyword density still an SEO signal?
Not as a target. Modern search engines use semantic analysis, and writing to a density percentage produces stilted text that reads as manipulation. Frequency analysis is now more useful for editing than for ranking.
What is Zipf's law?
The observation that the nth most common word appears about 1/n as often as the most common. It holds remarkably well across languages, so a frequency distribution that badly violates it is a sign of unusual or generated text.
How are non-English languages handled?
Tokenisation is the problem, not counting. Whitespace splitting fails on Chinese and Japanese, which need dictionary segmentation, and the stopword list is English-specific — so results for other languages need a matching list.
Should n-grams be counted as well as single words?
Often yes. Single words miss phrases — "machine learning" counted separately says less than the pair does. Bigrams and trigrams are what surface the actual subject matter of a document.
How should case and possessives be handled?
Lowercase before counting, or `The` and `the` split one entry into two. Possessives and plurals are harder — merging them needs stemming, which introduces its own errors, so most counters leave them separate and say so.

Common errors and gotchas

  • Leaving stop words in, which makes 'the' the top result and the analysis useless.
  • Counting a word and its plural separately, which understates the real emphasis.
  • Comparing counts between documents of very different lengths without normalising.
  • Ignoring case, or not ignoring it, without deciding which the question needs.
  • Treating frequency as importance, which it is only loosely related to.

Related Text Tools tools

Private & free — this tool runs entirely in your browser.