SMS character limits explained: 160, 153, 70 and 67

Why some texts use more credits than others. SMS character limits, GSM-7 and Unicode encoding, message parts, and how emojis and special characters affect cost.

Why character limits matter

Text messages have a fixed size. When your message is longer than that size, it is split into parts and reassembled on the recipient's phone, so they see one continuous message. On credit-based platforms, each part normally uses one credit. That means a message that is just a few characters too long can cost twice as much to send, and across a campaign to thousands of contacts the difference adds up quickly.

The number of characters that fit in each part depends on which characters you use. Most everyday English text fits within a standard character set, but a single emoji, curly apostrophe or unusual symbol can switch the whole message into a different encoding with a much smaller limit. Understanding the rules helps you write messages that are clear, affordable and predictable.

It also affects the reader. A short message that fits in one part is quicker to read and easier to act on than a long block of text. The limits are a useful discipline: they encourage you to say the important thing first and leave out anything that does not help the recipient. Most of the best business texts would fit comfortably in a single part.

GSM-7: the standard character set

Most texts are sent using an encoding called GSM-7. It covers the English alphabet in upper and lower case, numbers, common punctuation and some accented letters used in European languages. A single GSM-7 message can hold up to 160 characters. Spaces and line breaks count as characters, so a message with lots of blank lines uses up space faster than you might expect.

When a GSM-7 message is longer than 160 characters, it becomes a concatenated message. Each part needs a small amount of space for information that tells the phone how to join the parts back together, which reduces the limit to 153 characters per part. So a 161 character message uses two parts, a 307 character message uses three, and so on.

• 1 part: up to 160 characters.
• 2 parts: 161 to 306 characters.
• 3 parts: 307 to 459 characters.
• 4 parts: 460 to 612 characters.

Extended characters that count double

Some characters are part of GSM-7 but stored in an extended table, which means each one takes up the space of two characters. They are easy to use without noticing, especially when copying text from a website or document. The euro sign is a common example, as are several types of brackets and symbols used in web addresses and formatting.

• The euro sign.
• Square brackets and curly braces.
• The backslash, caret, tilde and vertical bar.

A handful of these in a message will not usually push it over a limit, but they can make the difference when you are close to 160 characters. If you are writing a long message, check the character count shown before sending rather than relying on a rough estimate. A message that sits at 158 characters can easily become 162 once a single euro sign or a pair of brackets is added.

Unicode: 70 and 67 characters

If a message contains any character that is not in GSM-7, the whole message is sent using Unicode, specifically an encoding called UCS-2. This supports almost every language and symbol, including emojis, but the limit drops sharply. A single Unicode message holds 70 characters, and when it is split into parts each one holds 67. A message that would fit in one GSM-7 part may need three Unicode parts.

Unicode is essential when you need to write in languages such as Arabic, Chinese, Polish with all its accents, or Urdu, and in those cases the extra cost is worth it to reach people in their own language. The problem is when Unicode is triggered by accident, by a character that adds nothing to the message and that you may not even realise is there.

Names and addresses can trigger Unicode too. If your contact list includes customers with accented names that fall outside GSM-7, a personalised message to those contacts may be sent as Unicode while the rest go as GSM-7. That is perfectly reasonable, as getting someone's name right matters, but it is worth knowing why the credit count for a bulk send may be slightly higher than you expected.

Emojis, curly quotes and other hidden culprits

Emojis are the most obvious trigger. A single smiley face switches the entire message to Unicode, and many emojis take up the space of two characters in that encoding. So a 150 character message with one emoji at the end can jump from one part to three. If you use emojis in marketing, do it deliberately and check the part count first.

Less obvious culprits come from copying and pasting text from word processors, websites or email. These programs often replace straight quotes and apostrophes with curly ones, turn hyphens into long dashes and add non-breaking spaces or ellipsis characters. All of these can switch a message to Unicode. Typing messages directly into the portal, or pasting into a plain text editor first, avoids most of these issues.

Writing efficient messages

The simplest way to control cost is to aim for a single part wherever possible. Most service messages, such as reminders, updates and codes, fit easily within 160 characters if you write plainly. For marketing messages, decide what is essential and cut the rest. A clear offer and an opt-out instruction will usually work better than a long description anyway.

• Write in plain GSM-7 text and avoid emojis unless they matter.
• Use straight quotes and apostrophes, and ordinary hyphens.
• Remove unnecessary line breaks and double spaces.
• Abbreviate dates and times sensibly, for example Tue 14 Oct, 10:20.
• Check the character and part count in the portal before sending.

Remember that merge fields, such as names or appointment times, vary in length between contacts. If your template is close to the limit, a long name or address could push some messages into a second part. Leaving a little room for variation keeps costs predictable across a bulk send, and saving tested templates means your team does not have to work this out each time.

Frequently asked questions