Word Counter

What actually gets counted

A word counter looks simple until two programs disagree by a few words on the same paragraph. The disagreement is almost always about definitions. This page uses these ones, so you can match them to whatever your teacher, editor or portal uses.

• Words are runs of letters or digits, split at spaces and punctuation. "Don't" is one word. "Well-known" is two, because the hyphen is treated as a break. "3.5" is one.
• Characters are what you see as single symbols, not raw code units. A Hindi syllable such as "क्षि" counts as one character, and an emoji counts as one, even though they are made of several code points underneath.
• Sentences end at full stops, question marks, exclamation marks, the Hindi danda (।) and the Urdu full stop (۔). A decimal point or an abbreviation in the middle of a line is handled by the browser's own sentence rules where available.
• Paragraphs are blocks separated by at least one blank line. A single line break does not start a new paragraph here, but it does add a line.
• Lines counts non-empty lines.

Why Hindi and Urdu text counts differently from a naive tool

The old way to count words is to split on spaces. That fails for languages that do not put spaces where English does, and it miscounts text with attached punctuation. A modern browser ships with a text segmenter that knows the Unicode rules for word, sentence and character boundaries, and this tool uses it when it is available. If your browser lacks it, the tool falls back to a pattern that matches letters, digits and combining marks, so Devanagari vowel signs stay attached to their consonants and do not split a word in half. The fallback is a little less precise for scripts that do not use spaces, such as Chinese or Thai, so treat counts for those scripts as close estimates.

A worked example

Consider this text: "Rain fell all night. By morning the lanes were flooded, and the school bus did not come."

• Words: Rain, fell, all, night, By, morning, the, lanes, were, flooded, and, the, school, bus, did, not, come. That is 17 words.
• Characters with spaces: 90. Without spaces: 74. The difference is the 16 spaces between 17 words.
• Sentences: 2.
• Reading time at 238 words a minute: 17 / 238 x 60 = 4.3 seconds, shown as 4 sec. Speaking time at 150 wpm: 17 / 150 x 60 = 6.8 seconds, shown as 7 sec.

Scale that up and the numbers get useful. A 1,000-word essay takes about 4 minutes 12 seconds to read silently (1,000 / 238 = 4.2 minutes) and 6 minutes 40 seconds to say aloud at 150 wpm. A five-minute speech at that pace is about 750 words, a number worth keeping in your head when you write one.

Where the 238 and 150 come from

The reading figure is the average silent reading speed for adults on non-fiction, about 238 words per minute, from a 2019 meta-analysis by Marc Brysbaert. Speaking pace in presentations is usually somewhere between 130 and 160 words a minute, so 150 is a middle value. Your own pace will differ. Dense technical text reads slower, and a nervous speaker with slides often goes faster than planned. Use the times as a planning figure: if the tool says 9 minutes and your slot is 10, rehearse once aloud, because the real number could be 8 or 11.

Using the repeated-word table

The table lists your most used words after removing common English filler words such as "the", "and" and "of". It is a quick editing aid. If "really" or "important" shows up eight times in 600 words, the paragraph probably leans on it. If a keyword you are writing about appears only once, you may have drifted off topic. It does not judge quality, and it will not catch repeated phrases or synonyms. Hindi and other languages have no stopword list here, so for those texts you will also see common small words like "है" in the table. That is expected, and you can read past them.

Target length and the common ways it goes wrong

Set a target in words or characters and the bar shows how close you are and how many remain. Characters are the right measure for things like SMS limits, meta descriptions or a form field that says "maximum 500 characters". Words are right for essays and articles. Watch for these traps:

• Counting the title, headings or references when the rule excludes them. Paste only the part that is being counted.
• Trusting a different tool's number. Word processors, portals and this page may disagree by a few words around hyphens, slashes and numbers. If a hard limit applies, leave a margin of 2 to 3 percent.
• Counting characters without spaces when the limit includes them. Check which version your form means.
• Pasting from a PDF. Line breaks from a PDF often split words with hyphens and inflate paragraph or line counts.

Limits

Everything runs in your browser and nothing is uploaded, so closing the tab erases your text. Very long documents of many lakh characters may feel slow on a phone. Words made of letters joined to emoji or unusual symbols may count differently in other tools. For text that must meet a legal or publishing limit, always verify in the system that will accept it. If you also need to keep a password, use something stronger than a document: the password generator on this site creates one without sending it anywhere.

Frequently asked questions