Word Counter

How to use the Word Counter

1. Paste or type your text. Counts, times and density tables update as you type — there is no button to press and no analysis to wait for. Deleting text updates everything just as immediately.
2. Read the counts you actually need. Word and character counts suit submission limits and platform rules. Sentence and paragraph counts, plus average words per sentence, suit structure checks. Characters without spaces suits the systems — some forms, some printing and messaging services — that measure a text without its spaces.
3. Use the times as averages, not promises. Reading time uses 200 words per minute and speaking time 130 words per minute. Your readers and your own delivery will be faster or slower; the figures exist so a script, a speech or a chapter can be estimated and compared, not so a stopwatch can prove them wrong.
4. Check the density tables for repetition. Scan the one-word table for a term you are leaning on too hard, then the two- and three-word tables for a phrase you repeat without noticing. Repetition that the tables expose in seconds is hard to feel while writing.
5. Copy the summary if you need the numbers elsewhere. The Copy counts summary button puts all eight headline figures in plain text, ready to paste into a document, a brief or a submission note.

Worked example

The sample text is two short paragraphs about tools and checklists — small enough to count by hand, structured enough to exercise every figure on this page. Select Load sample and the workspace fills with it. It contains a repeated phrase (good tools, twice), a blank line between its paragraphs, and three sentences of different lengths, which is exactly what makes the numbers below interesting rather than round.

Words: 22
Characters: 131
Characters without spaces: 109
Sentences: 3
Paragraphs: 2
Average words per sentence: 7.33
Reading time: 0.11 min (at 200 words per minute)
Speaking time: 0.17 min (at 130 words per minute)

Every figure can be checked against the rules stated on this page. The 22 words include the numbers-free but punctuation-carrying tokens such as errors, — a token counts because it contains letters, and the comma does not split it. The average of 7.33 is 22 words divided by 3 sentences. In the density tables, good leads the one-word phrases with 3 occurrences, which is 13.64% of the text’s words; good tools leads the two-word phrases with 2 occurrences, and because each occurrence accounts for two words, its percentage is 18.18% — a longer phrase can honestly outrank a single word, which is the point of measuring the share of words rather than raw occurrences alone.

What this tool counts, and what it deliberately does not

A word counter sounds too simple to argue about, until two counters disagree — and they do, because “a word” is a decision, not a fact of nature. This page makes its decisions in the open. A word is a whitespace-separated token containing at least one letter or digit, recognised in any script, so accented Latin, Arabic, Urdu and Chinese characters all count. A hyphenated word such as well-known is one token and counts as one word. A number counts as a word, decimals included as a single token. A token made only of punctuation — a dash standing alone, a row of asterisks — counts as nothing. An unspaced run of Chinese or Japanese characters is one whitespace token and therefore one “word” by this rule; that is a known simplification, stated here rather than hidden, and writers working in those scripts should treat the sentence and timing figures as rough.

The sentence count is the one figure built on a heuristic, and it is labelled approximate wherever it appears. Sentences are split on full stops, question marks, exclamation marks and ellipses, but a decimal point inside a number does not split, and a short list of common abbreviations — Dr., Mr., Mrs. and their kin — does not split either. That handles ordinary prose well. It cannot handle every name, initial or abbreviation in every language, and a sentence count for legal or literary-critical work should come from reading, not from this page.

Positioning matters, so here it is plainly: this tool counts and measures density only. It does not score readability — it will not tell you a reading grade, an ease score or whether your writing is “good”. The site’s Content Analyzer, which sits in the same AI & Content Productivity category as this tool, does readability: it applies published formulas to estimate how demanding a text is. Readers who need a grade go there. Readers who need exact counts — the student under a word limit, the speaker timing a script, the editor checking that a keyword has not been stuffed — come here, where every number can be traced back to a stated rule.

Common errors, and the limits of this tool

The commonest error is comparing counts between tools that count differently. A processor that treats hyphenated words as two, or numbers as non-words, will disagree with this page — not because one is broken, but because the rules differ. If a submission system gives its own count, that count wins for that submission; use this page for drafting and checking, and do a final check in the system that will judge you. The second error is reading density as a target. There is no percentage on this page that makes a text rank, convert or read well; the tables show what you repeated, and whether the repetition is a problem is a judgement only a reader of the whole piece can make. The third error is trusting the times for material with heavy numbers, formulas, code or lists — all of it counts as words here and none of it is read at prose speed.

The limits follow from the rules. Sentence counting is approximate, for the abbreviation reasons above. Reading and speaking times are averages at fixed rates — 200 and 130 words per minute — and say nothing about your audience or your delivery. Paragraphs are counted by blank lines, so a pasted document whose paragraphs are separated by single line breaks will read as one paragraph until blank lines are added. Density works on the exact tokens in the text: it lowercases and strips surrounding punctuation, but it does not stem words, so tool and tools are different phrases, and it does not merge synonyms. Finally, the analysis runs on what you paste — text locked inside images, scanned PDFs or tracked changes is invisible to it until it is extracted as plain text.

Frequently asked questions