Word frequency
Count repeated words, inspect the most frequent tokens, and download the complete word-count table as CSV.
How it works & useful details
How to use this tool
Paste your text to see the most frequent words. Set a minimum token length and optionally combine letter case. The table previews the first 200 rows; the CSV contains every result.
How the result is calculated
Words are runs of Unicode letters or numbers, with combining marks and internal straight or curly apostrophes. Hyphens split words. Text is NFC-normalized for counting; Chinese and Japanese runs are not dictionary-segmented.
Words are sorted by descending count, with ties in Unicode UTF-16 lexical order. The minimum length counts Unicode code points after case conversion. Case combining uses default Unicode lowercasing. Stop words such as “the” are included; this is not stemming or language-aware keyword analysis.
Up to 500,000 UTF-16 code units per input. Text stays on your device.
About this tool
Word frequency counts how often each word appears and lists the most frequent tokens, with a full table available as CSV. It is useful for content audits, keyword review, and spotting repeated phrases.
Tokenization is configurable so numbers, casing, and stop words can be handled the way you need. Analysis happens locally, and the CSV export contains the complete table.
How to use it
- Paste the text to analyze.
- Set the tokenization options, such as case folding.
- Review the most frequent words and their counts.
- Download the full frequency table as CSV.
Worked example
"The cat and the hat. The cat sat." with case combined produces:
the 3
cat 2
and 1
hat 1
sat 1That is 5 distinct words and 8 occurrences in total. Words with equal counts are listed alphabetically, so the order is stable across runs.
Making the results useful
Common function words such as "the", "and", and "of" dominate any text. Raise the minimum length to 4 or 5 to push them down and bring the content words forward. That is a quick way to see the themes in customer feedback, check whether a draft overuses a word, or compare the vocabulary of two versions of a document.
The count is exact, not language-aware: "run", "runs", and "running" are separate entries, since no stemming is applied. The CSV download contains every word, not just the first 200 shown in the preview, so you can continue in a spreadsheet.
Frequently asked questions
Are common words excluded?
No stop-word list is applied automatically; the full table includes every token so you can filter in your spreadsheet.
How are punctuation and numbers handled?
Tokenization splits on non-word characters by default, and the options let you refine what counts as a word.
What is the limit?
Up to 500,000 characters of text.