Word Count: Words, Characters and Reading Time (and Why the Numbers Differ)

Most text has three counts, and they answer different questions. Word count answers “how much have I written” — it is what essays, articles and reading-time estimates use. Character count answers “will this fit” — it is what form fields, meta descriptions, titles, tweets and SMS actually enforce. Characters without spaces is a third number that some platforms use, and it is the one most often confused with the other two.

The two are related but not interchangeable. English averages roughly five to six characters per word, so a 500-word article is about 3,000 characters — but “roughly” is doing real work in that sentence. Long words, URLs and code push the character count up without moving the word count, which is exactly the situation that makes people search for a counter in the first place.

Three things explain almost every counting surprise. Word counting is definition-dependent — hyphens and apostrophes change the answer. Character counting is space-sensitive — whether spaces count is a decision, not a fact. And limits are almost always measured in characters, even when they are described to you in words.

The rest of this article covers what each metric counts, why two tools disagree, the limits that matter in practice, how reading time is really calculated, and workflow habits for writing against a limit. If you just need a number, the word counter gives words, characters, characters without spaces and estimated reading time in one pass.

What Each Metric Actually Counts

The counts are mechanical, but only once you fix the definition. Here is what the standard definitions do to one test sentence:

The gaps between those three numbers are the whole story. The nine words spread across 45 characters, and only 36 of those characters are “content” — the rest is the whitespace that makes the text readable. That is nearly a 20% difference between two character counts of the same sentence, and it is why “characters” without a qualifier is ambiguous.

None of these definitions is more correct than the others. They measure different things, and the only error is not knowing which one your target platform is using.

Why Two Word Counters Disagree

Put the same text into two counters and you will often get two numbers, usually within a few percent of each other. The cause is not a bug — it is that “word” has no single universal definition when the text contains punctuation.

• Hyphenated terms. Is well-known one word or two? A whitespace-splitting counter says one; a counter that splits on hyphens says two.
• Apostrophes and contractions. It’s counts as one word under the common rule, but tools that treat apostrophes as word boundaries can read it as two fragments.
• Numbers, symbols and standalone punctuation. Some tools count 42 and & as words; others ignore anything without letters.
• Line breaks and multiple spaces. A tool that collapses whitespace handles these cleanly; one that naively splits on single spaces can overcount.

The practical consequence: for a school essay or a marketplace listing with a hard limit, the only count that matters is the one produced by whatever system will enforce the limit. Microsoft Word, Google Docs, a CMS field and a browser tool can each differ by a few words on the same document.

When the exact number matters, match the tool to the enforcer. When it does not, treat any reasonable counter as good enough — a 1% discrepancy in a 1,000-word draft is not worth the argument.

The Limits That Actually Constrain You

These are the numbers people are really checking against when they open a counter, and nearly all of them are character limits — including spaces:

The SEO limits deserve a word of caution, because they are not hard caps but soft truncation points. Google does not reject a 70-character title; it truncates or rewrites it in the results, which means the end of your title may never be seen. The practical goal is to get the important words within the first 60 characters, not to treat 60 as a cliff.

The SMS case is the one that surprises people most, and it is worth understanding because it costs real money. A standard segment holds 160 characters, but only when the text is plain ASCII under GSM-7 encoding. Add a single emoji or any non-Latin script and the message switches to UCS-2, which holds just 70 characters per segment. A message that was one segment becomes three.

How Reading Time Is Really Calculated

Reading time is an estimate built on one assumption: average reading speed. The standard figure used by most tools is 200 to 250 words per minute for an adult reading in their first language. The arithmetic is just division:

The honest caveat: this is a model, not a measurement. Real reading speed swings with the density of the text, the reader’s familiarity with the subject, and the language — and “reading” an instructional article with a tool open is far slower than skimming a news story. Two tools using 200 and 250 wpm will disagree by 20% on the same text, and both are defensible.

Treat reading time the way you treat a weather forecast: useful for planning, wrong in the specific instance. If a reader’s actual time on page matters to you, measure it directly rather than trusting the estimate.

Workflow Habits for Writing Against a Limit

The people who never fight their text fields do a few simple things consistently:

• Know which metric the limit uses before you write. Two minutes of checking whether a field counts words or characters, and whether spaces are included, saves a rewrite at the end.
• Write first, trim second. Getting a draft down and then cutting to the limit almost always produces better text than editing each sentence to length as you go.
• Front-load the important words. For titles and meta descriptions, the truncation point is what matters; put the keyword and the hook inside the first 60 and 155 characters respectively.
• Watch the long-content traps. URLs, code snippets and long unbroken strings inflate character counts far faster than word counts, and they are the usual reason a “short” post blows past a field limit.
• Keep a counter one paste away. The word counter shows words, characters with and without spaces, and reading time together, so you can check every relevant metric in one place. For the byte-level and encoding side of the same question, the string length tool counts bytes as well as characters.

None of this is complicated, which is the point. Counting is trivial once you know which number you are chasing; the failures come from guessing.

Count Your Text in One Place

Whatever limit you are writing against, the numbers should be instant:

• A free word counter with words, characters, characters without spaces, and estimated reading time.
• Runs entirely in your browser — nothing you paste is uploaded.
• Counts lines and paragraphs alongside words, for emails, posts and product descriptions.
• Part of the wider string tools set: character counter, case converter, URL encoder and Base64 encoder.

Frequently Asked Questions