# GSM-7 or Unicode: how an SMS’s length sets its price

> Why one character can double what a message costs, how SMS splits long messages into segments, and how to keep codes and alerts to one segment.

Source: https://sms.bweze.com/learn/sms-basics/gsm-7-vs-unicode/ · Updated 2026-11-14

An SMS is sent, and billed, in **segments**. BWEZE charges per segment, so the number of segments a message takes is what it costs. Two things decide it: the alphabet the message needs, and its length.

## Two alphabets

Every message is sent in one of two encodings.

| Encoding | Characters in one segment | Per segment when the message is longer |
| -------- | ------------------------- | -------------------------------------- |
| GSM-7    | 160                       | 153                                    |
| UCS-2    | 70                        | 67                                     |

**GSM-7** is the SMS default alphabet. It has the letters A to Z in both cases, the digits, common punctuation, and some accented letters, including é, è, à, ù, ì, ò, ä, ö, ñ and ü.

**UCS-2** is used as soon as one character in the message is not in GSM-7. The whole message switches, not just that character. Emoji and many accented letters do this. In French, ê, â, î, ô, û and ç are not in GSM-7, so one of them turns a 160-character message into a three-segment one.

## Why longer messages fit fewer characters

A message too long for one segment is split, and each part carries a small header that lets the phone put the parts back together. That header takes room. With GSM-7 a part holds 153 characters instead of 160; with UCS-2, 67 instead of 70.

So a 160-character GSM-7 message is one segment, but a 161-character one is two: 153 and 8.

## Characters that count twice

Ten GSM-7 characters live in an extension table and take two places each: `^ { } [ ] ~ \ | €` and the form feed. A message of 159 letters and one `€` is 161 places, so two segments.

BWEZE counts an emoji as two characters in UCS-2. It also never splits a two-place character across parts, so a long message can need one segment more than its length divided by 153 or 67 suggests.

## Examples

- `Your code is 482910` is 19 GSM-7 characters: **one segment**.
- A 200-character GSM-7 notice: 153 + 47, **two segments**.
- The same 100 characters with one emoji: UCS-2, 102 units, 67 + 35, **two segments**.

## Keeping a message to one segment

- Keep verification codes and alerts under 160 characters, without emoji.
- Watch for accented letters outside GSM-7 when you write in French; é and è are fine, ê and ç are not.
- Check a message before you send it. The SMS overview in the BWEZE dashboard counts a message's encoding and segments without sending it.

A single send can be at most 10 segments; the API refuses a longer one before it costs anything. The price of a segment is on the [SMS pricing page](/pricing/).

## Sources

- [3GPP TS 23.038: Alphabets and language-specific information](https://www.3gpp.org/DynaReport/23038.htm)
- [3GPP TS 23.040: Technical realization of the Short Message Service](https://www.3gpp.org/DynaReport/23040.htm)
- [BWEZE price list (GET /v1/pricing)](https://api.bweze.com/v1/pricing)
