Word Counter

Text:

Paste a draft, excerpt, caption, transcript, or drop one plain .txt file.

Drop TXT onto the textarea.

Case sensitive:

Off merges Apple and apple; on counts them separately.

Include numbers:

Turn off when numbers are clutter in prose review.

Remove stop words:

Use when repeated themes matter more than grammar words.

Normalize accents:

Use when sources mix accented and unaccented spellings.

Minimum word length:

Enter 1 or higher; use 2-3 to remove one-letter noise.

N-gram size:

Use 1 for word repeats, 2-3 for repeated phrasing.

N-grams across sentences:

Turn off when phrase repeats should restart at each sentence.

Word frequency rows:

Use 1 or higher; this changes only the exported frequency artifact size.

Repeated phrase rows:

Use 1 or higher; pair with N-gram size when checking repeated phrasing.

Metric	Value	Copy
{{ row.label }}	{{ row.value }}

Word	Count	Copy
{{ item.word }}	{{ item.count }}

N-gram	Count	Copy
{{ g.token }}	{{ g.count }}

Embed:

Customize

Include current inputs

Size

Advanced

Width

Height

Aspect ratio

Max height

Collapsible embed

Allow fullscreen

Referrer policy

Sandbox tokens

Length limits shape many kinds of writing before anyone judges the ideas. A school response may need to stay under a word cap, a caption may need to fit a narrow space, a grant abstract may need to prove focus, and a podcast script may need a realistic speaking length. Counting words is the first check, but it is rarely the only one that matters.

Text measurement starts with a choice that readers often do not see: what counts as a word. Contractions, hyphenated terms, numbers, accented letters, initials, bullets, and pasted table cells can all change the answer. Two editors can count the same draft differently because their token rules differ, not because either result is careless. Fair comparison means keeping the same rules each time.

Draft text separated into token rules, repeated patterns, and readability estimates

Word totals answer length questions. Sentence and paragraph counts show structure. Frequency lists reveal repeated words that may be intentional keywords, natural topic terms, or accidental clutter. Readability scores estimate effort from sentence length and syllable patterns, but they do not know whether the writing is accurate, useful, well organized, or right for its audience.

Common text measurement questions and the evidence that helps answer them
Question	Best evidence	Common mistake
Will the draft fit a limit?	Words, characters, lines, and estimated reading or speaking time.	Comparing two versions after changing token rules.
Does the prose feel repetitive?	Word frequency and repeated phrase counts.	Treating every repeated word as a problem, even when it is a necessary term.
Will a broad audience follow it?	Sentence length, readability scores, and a human review of purpose and audience.	Assuming a good score proves the message is clear.

How to Use This Tool:

Start with the text you want to measure, then decide whether you are checking a strict length limit, editing repetition, or comparing two drafts.

Paste text into Text, drag a plain TXT file onto the box, or choose Browse TXT for a longer draft.
Check the summary first. It shows Word Count, sentence and paragraph totals, estimated reading time, unique words, character count, readability band, lexical density, and the top word badge.
Open Advanced before comparing versions. Set Case sensitive, Include numbers, Remove stop words, Normalize accents, and Minimum word length once, then keep those settings unchanged for every draft in the comparison.
Use N-gram size and N-grams across sentences when repeated phrasing matters. Choose one word for simple frequency, two words for repeated phrase starts, or three words for longer boilerplate.
Review Text Statistics for the main counts and readability values, then move to Word Frequency and Repeated Phrases when the draft feels repetitive.
Use Word Frequency Chart, Word Length Chart, and Sentence Length Chart to spot concentration and outliers. Use JSON when you need the same measurements in a report or editing note.

If the sentence count, readability score, or phrase list looks surprising, check punctuation, abbreviations, decimal numbers, list bullets, and copied table cells before treating the result as an editing fact.

Interpreting Results:

Use the headline counts for limits and the deeper tables for editing judgment. A draft can meet a word cap and still be unclear, or score as readable while repeating a phrase too often for the intended audience.

How to interpret Word Counter outputs
Output	What to trust	What to verify
Words, characters, lines, paragraphs	Useful for length caps, form limits, script timing, and draft comparisons.	Token settings and pasted formatting can change totals.
Unique Words and Lexical Density	Shows vocabulary spread after the current token rules are applied.	A high percentage can come from a short or technical passage, not from better writing.
Flesch Reading Ease and Flesch-Kincaid Grade Level	Gives a rough difficulty signal from words per sentence and syllables per word.	Audience knowledge, layout, jargon definitions, and message order still need human review.
Word Frequency and Repeated Phrases	Flags repeated vocabulary, repeated openings, and possible keyword stuffing.	Common domain terms may need to repeat. Check whether the repetition helps or distracts.
Charts	Shows concentration, word-length spread, and sentence-length clusters at a glance.	Charts summarize the measured text; they do not decide whether the message is correct.

For readability bands, treat Easy read, Plain read, Dense read, and Very dense as editing prompts. They are not pass/fail grades.

Technical Details:

Text counting is a tokenization problem before it is a statistics problem. The same visible sentence can produce different totals when numbers are included, accents are folded, case is preserved, stop words are removed, or short tokens are filtered out. Those choices affect word count, unique word count, frequency rankings, repeated phrases, readability inputs, and histograms.

Sentence measurement uses punctuation boundaries, so periods, question marks, and exclamation marks carry more weight than line breaks. That is convenient for prose, but it can misread abbreviations, decimal numbers, bullet lists, initials, or pasted fragments. Paragraph count is based on blank-line separation, while line count follows line breaks in the pasted text.

Transformation Core:

How text is transformed into countable measurements
Stage	Rule	Effect on results
Character totals	All characters are counted, then whitespace is counted separately.	Separates form-limit checks from visible word totals.
Word tokens	Letter and mark sequences are counted as words, apostrophe forms are kept, and numeric tokens are counted when Include numbers is on.	Sets the base for Words, Unique Words, frequency lists, readability formulas, and charts.
Normalization	Case folding and accent normalization happen before grouping when those options are selected.	Determines whether variants such as Cafe and café group together.
Filtering	Stop words and words shorter than Minimum word length are removed from the measured token set when those controls apply.	Changes token-based counts and every derived value that depends on words.
N-grams	One-, two-, or three-token sequences are counted, either across the full text or restarted at each sentence.	Finds repeated single words, repeated phrase starts, and longer repeated wording.

Formula Core:

Readability formulas use average sentence length and average syllables per word. The syllable value is an estimate, so the score is best read as a warning signal rather than an exact measure of comprehension.

Average sentence length

ASL = \frac{words}{sentences}

Average syllables per word

ASW = \frac{syllables}{words}

Flesch Reading Ease

FRE = 206.835 - 1.015 ASL - 84.6 ASW

Flesch-Kincaid Grade Level

FKGL = 0.39 ASL + 11.8 ASW - 15.59

Lexical density

density = \frac{unique words}{words} \times 100

Reading and speaking time

minutes = \frac{words}{words per minute}

The reading estimate uses 200 words per minute; the speaking estimate uses 125 words per minute.

Readability Bands:

Flesch Reading Ease bands used by the Word Counter summary
Band	Flesch Reading Ease	Editing meaning
Easy read	80 or higher	Shorter sentences and simpler syllable patterns dominate the measured text.
Plain read	60 to 79.99	Often suitable for general prose, but still worth checking audience knowledge.
Dense read	30 to 59.99	Longer sentences or heavier vocabulary may slow readers down.
Very dense	Below 30	Review sentence length, terminology, and organization before relying on the text for broad readers.

Privacy Notes:

Text measurement runs in the browser. If you paste sensitive material, avoid sharing a page address until you confirm it does not contain the pasted text, because browser history and copied links can preserve address contents.

Worked Examples:

Short Safety Note

For "The safety checklist is ready. The safety checklist is ready. Review the ladder, gloves, and outlet before work starts.", Text Statistics reports 19 Words, 3 Sentences, and a Flesch Reading Ease of 62.38, which lands in Plain read. The repeated "safety checklist is ready" wording deserves an editorial check.

Repeated Phrase Review

For "Project notes repeat the launch plan. Project notes repeat the budget plan. Project notes repeat the launch plan.", set N-gram size to 3 and turn off cross-sentence grouping. Repeated Phrases shows "project notes repeat" and "notes repeat the" 3 times, plus "the launch plan" 2 times.

Decimal Number Check

For "Q2 revenue rose 12.5 percent. Q2 revenue rose 12.5 percent.", Include numbers increases Words from 8 to 10. The decimal point can also affect Sentences, so finance or science excerpts need a punctuation review before using readability scores.

FAQ:

Why does my word count differ from another editor?

Counters vary in how they treat contractions, numbers, punctuation, accented text, and pasted formatting. Use the same Advanced settings whenever you compare versions.

Do readability scores prove that the writing is clear?

No. Flesch Reading Ease and Flesch-Kincaid Grade Level estimate difficulty from sentence length and syllable patterns. Accuracy, organization, terminology, and reader knowledge still need review.

When should I remove stop words?

Turn on Remove stop words when topical repetition matters more than common English glue words. Leave it off when you need a normal word count or readability comparison.

Why do decimal numbers or abbreviations change sentence counts?

Sentence splitting depends on punctuation, so periods inside decimals, initials, and abbreviations can look like boundaries. Check Sentences before trusting readability scores for technical excerpts.

Glossary:

Token: A word-like unit after case, accent, number, stop-word, and length rules are applied.
N-gram: A sequence of one, two, or three tokens counted for repetition.
Lexical density: The percentage of measured words that are unique.
Flesch Reading Ease: A readability score based on average sentence length and average syllables per word.
Flesch-Kincaid Grade Level: A readability estimate that maps the same sentence and syllable inputs to a grade-level style number.

References:

Measuring Readability, Indian Health Service.
Clear Communication Index User Guide, Centers for Disease Control and Prevention, last reviewed November 19, 2020.
Plain Language Materials & Resources, Centers for Disease Control and Prevention, July 21, 2025.