Paste or type. Stats update instantly.
Count words, characters, sentences, paragraphs, and reading time in any text. Get instant statistics as you type, without sending your text anywhere. The Word Counter handles essays, blog posts, social media drafts, academic papers, and code comments with the same accuracy. Everything runs locally in your browser.
"Word count" sounds like a single number, but a useful word counter exposes several related metrics that answer different questions. This tool reports six.
Word count. How many words your text contains. Words are detected as runs of non-whitespace characters, with hyphenated compounds like "self-driving" counted as one word and contractions like "don't" counted as one word. This matches Microsoft Word's count, which is the de facto standard for academic and professional contexts.
Character count (with spaces). Total characters including every space, tab, and newline. This is what social platforms use for their limits. Twitter's 280, SMS's 160, meta description's 155.
Character count (without spaces). Same total minus whitespace. Used in some printing and typesetting contexts; rarely the right number for digital constraints, but commonly requested for academic submissions in some countries.
Sentence count. Sentences detected by terminal punctuation (periods, question marks, exclamation marks). Abbreviations like "Dr." and "e.g." can fool the simple detection; the tool tries to handle the common ones.
Paragraph count. Blocks of text separated by blank lines. Single-line breaks within a paragraph aren't counted as paragraph boundaries.
Reading time. Estimated based on an average reading speed of 200 words per minute for non-fiction text. The number is a rough guide, not a precise prediction. Technical content reads slower (100β150 wpm); fiction reads faster (250β300 wpm).
Hitting word count requirements for academic papers. Essays, theses, journal submissions, and grant applications all enforce strict word limits. The Word Counter shows your count as you type, so you can adjust before hitting the cap or finishing short.
Optimizing blog post length for SEO. Long-form blog posts (1,500β2,500 words) consistently rank better than short ones in Google, but only if the content earns the length. The counter helps you hit a target without padding for its own sake.
Fitting social media character limits. Twitter (280), LinkedIn (3,000 visible, 700 recommended), Instagram captions (2,200), Mastodon (500). Different platforms, different ceilings. Character count with spaces is what each platform actually counts.
Meta description and title tag optimization. Search engines truncate meta descriptions around 155β160 characters and titles around 60. Counting precisely lets you write right up to the limit without overflow.
Estimating reading time for newsletters and articles. "5 min read" labels tell readers what they're committing to. Reading time at 200 words per minute is the common standard.
Speech and presentation timing. Speakers deliver around 150 words per minute conversationally; a 20-minute keynote needs about 3,000 words of script. Word count gets the timing right.
Cover letter and resume length checks. Cover letters typically run 250β400 words; resumes peak around 500 words for early career and 700 for senior. Counting prevents both verbose padding and underdone first drafts.
Email subject line optimization. Subject lines under 50 characters perform better in most email clients (which truncate at varying lengths around 50β60). The character counter applies here too.
Different tools count differently, and the differences matter for academic submissions where "exactly 2,000 words" matters.
Hyphenated compounds. "Self-driving" is one word in Microsoft Word, Google Docs, and this tool. Some older counters split on hyphens and call it two words. The modern convention is to count compounds as single words, matching how readers perceive them.
Numbers and abbreviations. "1,000" is one word. "U.S.A." is one word. "Dr. Smith" is two words. The counter treats any run of non-whitespace as a word, which produces the right result for these cases.
URLs and email addresses. A URL like "https://example.com/path" is one word. An email like "name@example.com" is one word. If you need URLs excluded from the count (academic submissions sometimes exclude footnotes and bibliographic URLs), strip them manually first.
Smart quotes and special characters. Curly quotes, em dashes, en dashes, and other typographic characters don't affect the word count. They're just attached to adjacent words.
Empty lines and trailing whitespace. Don't affect word count. They affect character count if you include spaces.
This tool. Instant, in-browser, includes reading time and paragraph count by default. Best for quick checks and for users not in a word processor.
Microsoft Word's count (Tools β Word Count, or the bottom status bar). Authoritative for academic and professional documents that specify "Word count per Microsoft Word." Slight differences from this tool are normal; the same input typically differs by 1β3 words.
Google Docs's count (Tools β Word count, or Ctrl+Shift+C). Matches Microsoft Word's algorithm closely. Useful for collaborative documents.
For documents where exact count matters (a grant application capped at exactly 5,000 words), use the tool that's specified by the submission requirements. For everything else, this tool is faster.
Counting placeholder text. If you have Lorem Ipsum or notes-to-self in the document, the counter includes them. Remove placeholder text before measuring against a real word limit.
Forgetting URLs and footnotes count. Academic guidelines often specify that bibliographic references and footnotes don't count against the word limit. The counter doesn't know about these conventions; you'll need to manually subtract them or paste only the body text.
Pasting from PDF or HTML. Copy-pasting from PDFs sometimes produces double-spaces or stray characters that throw off the character count slightly. Run the input through Remove Spaces (collapse mode) to normalize before counting precisely.
Different reading speeds for different content. The 200 wpm reading time estimate fits typical non-fiction prose. Technical content with heavy jargon reads slower; light entertainment reads faster. Don't take the estimate as gospel for niche content.
Type or paste text into the input box. The counter updates live as you type. Every keystroke produces a new count, with no debouncing for typical document sizes. Detection uses standard regex patterns: words are runs of \S (non-whitespace), sentences are terminal-punctuation-bounded segments, paragraphs are blocks separated by double newlines.
The reading time formula is simple: words / 200, rounded to the nearest minute. This matches the standard convention used by Medium, Substack, and most editorial platforms.
Performance scales linearly with input size. Multi-megabyte documents (hundreds of thousands of words) count in under a second on modern browsers.
Set a target before you start writing. Decide on the word count goal before drafting. Writing toward a known target produces tighter output than writing freely and trimming later.
For SEO, aim for 1,500β2,500 words on cornerstone content. HubSpot's analysis of 6,000 articles found 2,100β2,400 words as the sweet spot for organic traffic. Below 1,500 you compete with thin content; above 2,500 you risk losing readers.
For social posts, write to fit the platform. Twitter posts under 100 characters get more engagement than longer ones. LinkedIn posts perform best between 150β500 characters in the visible portion. Instagram captions can be longer (1,000+) without penalty if structured well.
Use reading time as a planning tool. If your post is 8 minutes and you're aiming for 5, decide what to cut. If it's 3 and you're aiming for 7, decide what to expand.
One. "Well-known" is one word, matching Microsoft Word and Google Docs. Some older counters split on hyphens; the modern convention is single-word counting.
Approximate. Based on 200 words per minute, which matches average non-fiction reading speed. Faster readers and lighter content read 250β300 wpm; technical readers and dense content read 100β150 wpm. Take it as a guide.
For space-separated languages (English, Spanish, French, German, Russian, etc.), yes. Accurately. For Chinese, Japanese, and other scripts that don't use word separators, the count won't be meaningful since the tool detects words by whitespace breaks. Character count works for any language.
Yes. The tool strips formatting on paste in most browsers, leaving just the text. Tables, images, and inline formatting are removed; only the actual text content is counted.
No. Counting runs entirely in your browser. Nothing is uploaded, logged, or stored. You can verify by opening your browser's network tab and watching it stay empty as you type.
Different tools handle edge cases differently. Abbreviations, embedded numbers, hyphenated terms. Differences of 1β3 words on a 1,000-word document are normal. For documents with strict limits, use the counter the submission requires.