Characters With Spaces vs. Without: What Counts as a Space, and Who Counts What

Almost every character counter shows two numbers: characters with spaces and characters without. It is easy to assume the difference is just the number of space bars you pressed. It isn't, because the phrase "without spaces" has no single definition. Tools disagree about whether a tab, a non-breaking space or a line break is a space, and some professions fixed their own convention long before online counters existed. This post covers which convention applies where, and what the tools actually do. For the basics, see Word Count vs Character Count and Character Count in Google Docs.

Check your own text with Clear Count, which counts words and characters in your browser. Nothing is uploaded.

Which Convention Applies Where

Translation is the clearest example of two conventions living side by side. German translation agency Intercontact describes a standard line as 40 characters excluding spaces, or 50 to 55 including spaces, and says to divide a text's character count with spaces by 55 to get its line count. A standard page is given two ways: 25 lines of about 40 characters, roughly 1,000 characters without spaces, or 30 lines of up to 60 keystrokes, a maximum of about 1,800 characters with spaces. The agency stresses that the agreed definition must be fixed in advance.

The last row is a rule of thumb that our Google Docs post also gives: the with-spaces figure is usually the relevant one for external limits. Always confirm with the platform's own documentation.

What Counts as a Space

I tested one short string: "Hello world", a tab, "foo", a non-breaking space, "bar", a line break, "baz", two ordinary spaces, and "qux". It has 28 characters. Removing "spaces" gives different answers depending on the definition. I ran this in Node 22, and Python 3's isspace() gave the same 22 for the last row.

"Characters without spaces" sounds like a precise measurement. On a single line it ranged from 22 to 25, a difference of about 14 percent, depending only on what the tool decided a space was.

Not every invisible character counts as whitespace either. A zero-width space stays in the count in both languages, which is one reason pasted text can look shorter than it measures. We cover that in The Invisible Characters Hiding in Text You Copy and Paste.

Real Tools Disagree Too

An UltraEdit support thread explains why its count differed from Microsoft Word's. UltraEdit treats any Unicode whitespace character as a space. Word's "with spaces" figure, according to that thread, covers ordinary spaces, non-breaking spaces and tabs but not carriage returns or line feeds, and treated an EN SPACE as a non-space character in Word 2007. In their test file the whitespace totals were 3,208 and 2,930. The thread concerns an old Word version, so current behavior may differ.

How to Avoid a Wrong Number

• Ask which definition is meant. "Without spaces" in a translation quote may not match your word processor's figure.
• Count in the destination when a limit is tight. A text pasted from Word may contain tabs and non-breaking spaces that another tool counts differently.
• Check line breaks. Some counters ignore them and some include each one as a character.
• Leave a margin. If the limit is 280 and your count is 279, assume you may be over.

---

The honest summary: "with spaces" is the safer number for platform limits, and "without spaces" is mostly a legacy convention from translation and publishing. The without-spaces figure is not one number but at least four, and in my test one short line scored between 22 and 25. Agree on the definition, or count with the tool that will make the final check.