Multilingual SMS

Send messages in your customers' preferred language to increase engagement and build trust. When you operate across regions with multiple languages, multilingual SMS ensures your communication remains consistent. In regulated markets, localized communication may be required by law. Infobip SMS supports a wide range of encodings to ensure your messages are delivered correctly across different languages and regions.

Technical considerations [#technical-considerations]

Sending SMS in multiple languages introduces technical considerations:

• Character sets: Different languages require different characters. While the GSM 7-bit alphabet covers basic Latin characters, many languages require Unicode encoding.
• Encoding: Using Unicode (UCS-2) reduces the maximum characters per message (from 160 to 70), which can lead to concatenation and increased costs.
• Unicode compatibility: Some carriers and devices may not support all Unicode character ranges, which can affect message delivery or cause character substitution.

Character encoding in SMS [#character-encoding]

SMS messages can be sent in different character encodings depending on the language and characters used. The choice of encoding affects message length, concatenation, delivery, and cost.

Supported languages and regions [#supported-languages-and-regions-character-encoding]

The following table shows which languages and regions each encoding typically covers.

GSM 7-bit alphabet [#gsm-7bit-character-encoding]

The GSM 7-bit alphabet is the default encoding for SMS.

Supported characters: Basic Latin letters (A-Z, a-z), numbers, and common punctuation. Extended characters (such as ^, {, }, €, [ ]) use an escape mechanism that reduces message length.

Typical languages: English, Spanish (without accents), German, Italian, and many Western European languages.

Message length: Up to 160 characters per single-part SMS. When extended characters are included, fewer than 160 characters may fit.

Basic characters:

Extended characters (require escape sequence, reduce message length):

Concatenated GSM messages [#concatenated-gsm-character-encoding]

If a message is longer than 160 characters, it is split into multiple parts (concatenation). Each part can contain up to 153 characters because additional bytes are reserved for headers that reassemble the message on the recipient's device.

Unicode (UCS-2) [#unicode-character-encoding]

Unicode encoding is used when characters outside the GSM 7-bit alphabet are included in the message.

Concatenated Unicode messages [#concatenated-unicode-character-encoding]

If a Unicode message is longer than 70 characters, it is split into multiple parts. Each part can contain up to 67 characters because of the concatenation header.

Delivery and cost implications [#delivery-cost-character-encoding]

Encoding choice affects message length, delivery, and cost in the following ways.

Reducing encoding impact [#reducing-encoding-impact]

When a message includes any character not supported by GSM-7, the entire message is encoded in Unicode, reducing the maximum length to 70 characters. This single non-GSM character forces the whole message into Unicode encoding.

To reduce encoding impact and preserve standard message length, Infobip offers two approaches:

National Language Identifier [#national-language-identifier-reducing-encoding-impact]

National Language Identifier (NLI) is an encoding technology in the GSM standard that allows SMS messages to identify the language of the content. It indicates that the message contains language-specific characters that must be delivered as the original text, without defaulting to full Unicode.

NLI-encoded messages can contain up to 155 characters (five characters are reserved for metadata that informs the receiver's device about the selected language and how to display the SMS).

Supported languages:

• Turkish (TR)
• Portuguese (PT)
• Spanish (ES)
• AUTODETECT: Automatically detects the language from message content. Detection is limited to Turkish, Portuguese, and Spanish. If no supported language is detected, the message falls back to standard encoding.

Turkish supported characters:

Portuguese supported characters:

Spanish supported characters:

You can configure NLI through the API or the web interface. In the API, include the languageCode parameter in your SMS request. In the web interface, select the appropriate NLI character set (Turkish NLI, Portuguese NLI, or Spanish NLI). See Send SMS message for API details.

Important notes:

• Some networks might not support the Language feature. This functionality may not work for all destinations. For example, a Turkish message sent through a Chinese provider might not display correctly on the recipient's device.
• Non-standard characters might cause messages to be encoded in Unicode, reducing available characters.
• Always test messages to confirm encoding.

Use NLI when sending messages in Turkish, Portuguese, or Spanish with language-specific characters and you need more than 70 characters.

SMS transliteration [#sms-transliteration-reducing-encoding-impact]

SMS transliteration converts text written in a native script into a Latin-script version of the same message. It replaces any special (unsupported) characters with similar GSM-7-compatible ones, ensuring that the maximum of 160 characters can be used instead of only 70.

Supported languages:

• Turkish
• Greek
• Cyrillic
• Serbian Cyrillic
• Bulgarian Cyrillic
• Central European
• Portuguese
• Colombian
• Baltic
• NON_UNICODE

When you specify the desired output alphabet, unsupported characters are converted differently depending on which character is most appropriate for the selected language. Any character that is not recognized by the selected language and is not part of the GSM alphabet is replaced by a dot (.).

Greek supported characters for transliteration:

Cyrillic supported characters for transliteration:

Turkish transliteration character mapping:

NON_UNICODE transliteration: Converts any Unicode text to GSM 7-bit alphabet by mapping characters to closest GSM equivalent. Unmappable characters become dots (.).

Other language transliterations (Portuguese, Colombian, Baltic, Serbian Cyrillic, Bulgarian Cyrillic, Central European) follow similar character mapping patterns to preserve readability while staying within GSM 7-bit constraints.

Example NON_UNICODE transliteration:

Example Greek transliteration:

By using transliteration, Greek lowercase letters that are not supported in the GSM alphabet are converted to uppercase letters that are supported. The message remains readable while preserving the 160-character limit.

You can configure transliteration through the API or the web interface. In the API, include the transliteration parameter with the desired output alphabet (TURKISH, GREEK, CYRILLIC, and more). In the web interface, select the appropriate character set from the available options. See Send SMS message for API details.

Use transliteration when sending messages in Greek, Cyrillic, Turkish, or other supported languages and you need the full 160-character limit without splitting into Unicode parts.