How each number is defined
Word and character counting sounds trivial until two tools disagree, so here are the exact definitions used. They aim to match what a careful human editor would count, not what a byte-length function returns.
| Metric | Rule | Example |
|---|---|---|
| Words | A run of letters or digits, optionally joined by an inner apostrophe, full stop, hyphen or underscore. | don't and e-mail are one word each; -- is none. |
| Characters with spaces | Every user-perceived character, including spaces, tabs and line breaks. | a b → 3 |
| Characters without spaces | The same count with all whitespace removed first. | a b → 2 |
| Lines | The number of lines in the box. A trailing newline does not create an extra line. | one\ntwo → 2, and so does one\ntwo\n. |
| Paragraphs | Runs of consecutive non-blank lines, separated by one or more blank lines. | Three blocks of text with blank lines between them → 3. |
| Sentences | Text ending in ., !, ? or …, ignoring fragments with no letters or digits. |
Hi. Bye! → 2 |
| Reading time | Words ÷ 200 words per minute, the usual average for adult silent reading. | 400 words → 2 min |
| Longest line | Character count of the longest single line, useful for spotting lines that will overflow. | — |
Why character counts differ between tools
Most counters report JavaScript string length, which counts UTF-16 code units. That makes “😀” count as two, and a family emoji such as “👨👩👧” count as eight or more. This tool counts grapheme clusters — what you actually see — so an emoji is one character. The secondary line under the counters also shows the Unicode code-point count and the UTF-16 unit count, so you can see all three figures when a platform limit depends on a specific one.
Practical uses
- Checking an article against an editorial word budget before filing.
- Fitting social posts, meta descriptions and ad copy to a character limit.
- Verifying an essay or assignment against a minimum and maximum length.
- Estimating transcription, voice-over or reading time from a script.
- Sanity-checking a translation against its source, where wildly different counts usually mean missing text.
Counts never leave your browser. There is no “analyse” request and no server-side word counting, so you can safely check confidential drafts, contracts or client material.
Frequently asked questions
- Do the counts include the text I deleted?
- No. Only the current contents of the box are measured.
- Why does my word processor report a different word count?
- Word processors usually count hyphenated words as one but treat some punctuation differently, and they count headers, footers and footnotes that are not in this box. For plain body text the two agree closely.
- Does downloading the text store it anywhere?
- No. The download is generated in the browser with a temporary object URL and saved straight to your own machine.
- How many words is a 5 minute read?
- About 1,000 words. The reading time here uses 200 words per minute, the usual average for adult silent reading, so 5 minutes is roughly 1,000 words and a 1,000 word article reads in about 5 minutes.
- How many words is a 1,000 character limit?
- Roughly 150 to 200 words in English, because the average English word is about 5 characters plus a space. Paste your text above and the character counts update as you type, so you can check a real limit exactly.
- How do I count characters without spaces?
- Read the Characters without spaces figure, which removes all whitespace before counting. Many social platforms and meta description limits are defined that way, so it is worth checking both numbers.
- Do emoji count as one character or two?
- One, here. Most counters report JavaScript string length, which counts UTF-16 code units and makes an emoji count as two. This tool counts the grapheme clusters you actually see, and shows the code-point and UTF-16 figures underneath so you can check a platform-specific limit.