SMS Encoding: GSM-7 & Unicode/UCS-2

Overview

There are many character sets for SMS.  This article will focus on the 2 most common and widely used.  While Voxox supports many encodings, if you are troubleshooting content delivery issues or new to SMS we suggest that you start with GSM-7 and Unicode.

In most cases, GSM-7 will be all you need as it supports all Latin characters A-Z and digits 0-9 as well as some special characters.

If you are using GSM-7 (or not specifying a character set at all) and your messages are received with odd symbols or ?, try setting your encoding to to Unicode/UCS-2. 

GSM-7

For more detail on the GSM-7, here is a link to the character set for GSM-7.  If your messages are being delivered with odd symbols or ?'s where there should be a character, its very likely that it does not appear on this character set.   A single GSM-7 message will allow 160 characters.   (see below for longer multi-part messages)

Unicode

SMS sent and received using characters not found in the GSM-7 character set require the use of Unicode or UCS-2 encoding.

Unicode will handle any character you throw at it, but that the cost of the characters allowed per message segment.  A single Unicode SMS message will allow 70 characters.  (see below for longer multi-part messages)

Hint: Don't know which method to use? Try entering your text into this tool by Chad Selph. If its red, try changing to UCS-2.

Multipart messages

When a GSM-7 message requires more than 160 characters, it will be split into multiple message parts and automatically recombined on the user handset.

In order for the message to be reconstructed correctly on the handset, a User Data Header (UDH) will be added to each message.  This extra data will reduce the per segment size to 153 GSM-7 characters per segment.

In the case of UCS-2 characters, a multi-part message will allow 67 UCS-2 characters per segment.