Text tool

Word & Character Counter

Counts update as you type. Alongside the usual totals you get reading time, speaking time, and a frequency table showing which words you have leaned on too hard.

Your text

0Words

0Characters

0No spaces

0Sentences

0Paragraphs

0Lines

Time on the page

0 sec

Read silently at 238 words per minute · spoken aloud at 140.

Why two counters disagree about the same text

Paste one paragraph into three different tools and you can get three different word counts. None of them is broken — they are answering slightly different questions.

Hyphenated words. Is state-of-the-art one word or four? Splitting on whitespace, as this tool does, gives one. Microsoft Word agrees. Some SEO tools split on hyphens and give four.
Numbers and symbols. $1,200 counts as one word here. Tools that strip punctuation first may count it as zero.
Em dashes with no spaces. later—much later is one word when splitting on whitespace, two when splitting on punctuation.

This tool splits on whitespace and nothing else, which is the same rule word processors use. If you are writing to a limit set by a publisher, that is almost always the rule they meant.

Where the reading-time figure comes from

Reading speed here is set at 238 words per minute for silent reading and 140 for reading aloud. The silent figure comes from research reviewing reading-rate studies across many experiments; it is a mean for adults reading non-fiction English, not a target. Real speed varies a lot with the material.

Roughly how long a piece takes to read silently at 238 wpm
Words	Typical piece	Reading time
50	Product description	13 sec
150	Press release intro	38 sec
500	Short blog post	2 min 6 sec
1,000	Standard article	4 min 12 sec
2,500	Long feature	10 min 30 sec
10,000	Dissertation chapter	42 min

Dense technical writing runs slower — closer to 150 wpm once you allow for re-reading. Speaking is slower still, which is why a 1,000-word script fills about seven minutes of video, not four.

Character limits worth knowing

Character counts matter more than word counts when you are writing into a box with a hard cap.

Character limits — the SMS figures are a fixed technical standard, the rest are platform settings that can change
Where	Limit	Detail
SMS, one segment	160	Using the GSM-7 alphabet. A fixed standard, not a policy.
SMS with any emoji or Urdu	70	One non-GSM character switches the whole message to UCS-2 and more than halves the limit.
SMS, multipart	153 each	Seven characters per segment go to the joining header.
Meta description	~155	Google truncates on pixel width, not characters — see below.
Title tag	~60	Also pixel-based. Wide capitals eat the budget faster.

The SEO ones deserve a warning. Search engines cut snippets by pixel width, so a title of sixty narrow characters (illiterate militia) fits where sixty wide ones (WOMAN AWARDED MMXXV) does not. Treat 60 and 155 as rough budgets, then check how the result actually renders.

Reading the frequency table

The frequency table skips roughly forty filler words — the, and, of, to, is and their relatives — because they top every list and tell you nothing. What is left is the vocabulary you actually chose.

A single content word above about 3% of your total usually reads as repetitive to a human, and that threshold matters more than any SEO rule of thumb. If solution appears 34 times in 900 words, the fix is not a lower percentage — it is finding out that you meant three different things by it and naming them separately.

Questions people ask

How are sentences counted?

By splitting on full stops, question marks and exclamation marks followed by a space or the end of the text. Abbreviations like e.g. and Dr. will inflate the count slightly — no counter solves this perfectly without a dictionary, and this one does not pretend to.

What counts as a paragraph?

A block of text separated by one or more blank lines. Single line breaks inside a block do not start a new paragraph, which matches how most editors behave.

Does it handle Urdu, Arabic or other right-to-left text?

Word and character counts work correctly, since both are based on whitespace and Unicode code points. Sentence detection is weaker, because Urdu uses the full stop character ۔ rather than a period — that character is included in the split, but mixed-script text can still confuse it.

Is there a size limit?

No fixed one. Counting is fast up to a few hundred thousand words on a laptop; on an older phone you may notice a pause past about 50,000. The frequency table is the slow part, so it only rebuilds when you stop typing.

Is my text uploaded?

No. Everything happens in this tab. There is no server call, no storage, and closing the page erases it.