Character Counter

Updated September 2026

Why SMS is billed by segment, not by message

This is the part almost no character counter gets right, and it is the one with a bill attached.

A text message is not sent as characters. It is sent as a fixed number of bits, and how many of your characters fit depends on the alphabet the carrier has to use. Plain Latin text is encoded in GSM-7, a 7-bit alphabet that fits 160 characters in one segment. The moment your text contains one character outside that alphabet, the whole message is re-encoded in UCS-2, which is 16 bits per character and fits only 70.

It gets worse when a message is long enough to split. Concatenated messages spend part of each segment on a header telling the phone how to reassemble them, so the per-segment capacity drops:

The practical consequence: a 300-character marketing text in plain English is 2 segments. Put one emoji in it and it becomes 5. On a campaign to 50,000 subscribers that is the difference between 100,000 and 250,000 billable messages. The counter above shows the encoding and the segment count live, so you can watch a single character change the bill.

Two characters catch people out constantly. The curly apostrophe that word processors insert automatically is not in GSM-7 — the straight one is. Same with the em dash. Copy a sentence out of Word into an SMS tool and you can trip into UCS-2 without typing anything unusual at all. Six GSM characters also cost two units each rather than one: ^ { } \ [ ] ~ | and the euro sign.

Not every accent breaks GSM-7 — and that surprises people

The GSM alphabet includes a handful of accented characters, but not the ones you would expect. è, é, ù, ì, ò, à, ä, ö, ñ and ü all survive in GSM-7. But á, í, ó and ú do not, and neither does any capital accented letter beyond Ä, Ö, Ñ, Ü and É.

For anyone sending Spanish-language SMS this is a live trap. "Café" costs you nothing extra. "Mamá" halves your capacity. The same word written "mama" without the accent stays in GSM-7 and fits in a single segment. The counter above tells you which side of that line your text is on before you send it.

Six more characters are in GSM-7 but cost two units instead of one, because they live in an extension table the phone has to be told to switch into: ^ { } \ [ ] ~ | and the euro sign €. A message full of square brackets runs out of room sooner than its character count suggests.

What a segment actually costs

Bulk SMS is priced per segment, not per message. At a typical U.S. rate of roughly one cent per segment, the arithmetic on a campaign looks like this:

One emoji added $1,500. Cutting forty characters saved $500. This is why the segment count, not the character count, is the number that matters when you are writing an SMS campaign.

Why two tools give different counts

If this counter disagrees with Word or Google Docs by a few characters, one of these is usually why:

• Line breaks. A line break is a character. On Windows it is often two — carriage return plus line feed — so text pasted from a Windows editor can count higher than the same text typed in a browser.
• Trailing spaces. A space at the end of a line is invisible and counted. Several at the end of a paragraph add up.
• Word's autocorrect. Word silently replaces straight quotes with curly ones and double hyphens with em dashes. The character count barely moves, but for SMS it changes the encoding entirely.
• Emoji. Some tools count an emoji as one character, others as two, and emoji with skin-tone or gender modifiers can count as four or more. There is no single right answer — what matters is what the destination platform counts.

Every limit that matters, in one place

The panel above tracks these live. Here they are for reference, with what actually happens when you go over:

Note the difference between a hard limit and a display limit. Instagram will happily accept 2,200 characters, but only about 125 are visible before the caption collapses — so the sentence that has to land goes first. Google will index a 300-character meta description and simply show 155 of it. Writing to the display limit rather than the hard limit is usually the better instinct.

Title tags are measured in pixels, not characters

Google does not truncate search titles at a character count. It truncates at roughly 580 pixels of rendered width. A title full of narrow letters like "i" and "l" survives longer than one full of "W" and "M". Sixty characters is a safe working rule, but if your title is packed with wide capitals, assume less.

With spaces or without?

Almost every platform limit counts spaces, which is why the big number above is the one to watch. The "no spaces" count exists for two real cases: translation and transcription quotes, which are frequently billed per character excluding spaces, and a handful of academic submission rules that measure visible characters only. If nobody told you explicitly to exclude spaces, include them.

Frequently asked questions