SMS Encoding: GSM-7 vs UCS-2 Explained

Every SMS must be encoded before transmission. The encoding determines how many characters fit in a single SMS segment (160 or 70 characters) and therefore how many segments are needed — which directly affects cost. There are two primary encodings: GSM-7 and UCS-2.

GSM-7 (also called GSM 03.38) is the default encoding for SMS. It uses 7 bits per character, allowing 128 basic characters plus a 10-character extended table. A single GSM-7 SMS segment holds 160 characters. For multipart messages, each segment holds 153 characters (7 characters are used for the User Data Header). The GSM-7 character set includes: uppercase and lowercase Latin letters (A–Z, a–z), digits (0–9), common punctuation, and some special characters like @, £, $, ¥, è, é, ù, ì, ò, Ç, Ø, ø, Å, å, Δ, _, Φ, Γ, Λ, Ω, Π, Ψ, Σ, Θ, Ξ. Extended GSM-7 characters (escaped, count as 2 chars): [, ], {, }, \, ^, ~, |, €. Using any character NOT in the GSM-7 set will trigger UCS-2 encoding for the entire message.

UCS-2 (Universal Character Set — 2 bytes per character) supports the full Unicode Basic Multilingual Plane. This allows sending Arabic, Chinese, Japanese, Korean, Cyrillic, Greek, Hindi (Devanagari), emoji, and thousands of other characters. The tradeoff: 2 bytes per character means only 70 characters per segment (or 67 per segment in multipart messages). This nearly halves the characters-per-pound efficiency and roughly doubles the number of segments needed for equivalent content. UCS-2 is triggered automatically by our API when your message content contains any non-GSM-7 character. You don't need to specify encoding — it's detected and applied transparently.

These characters look harmless but are NOT in GSM-7 and will force UCS-2 encoding — doubling your segment count and cost:

The BulkSMSRates API automatically detects the required encoding and reports it back in the API response. You can also request specific encoding.

POST /v1/send
{
  "to": "+447700900000",
  "from": "MyBrand",
  "body": "Hello! Your order is ready 🎉"
}

// Response:
{
  "data": {
    "message_id": "msg_abc123",
    "encoding": "UCS-2",     // 🎉 triggered UCS-2
    "segments": 1,
    "body_length": 30,
    "chars_remaining": 40,
    "cost": 0.0300,
    "currency": "GBP"
  }
}

1. Avoid smart quotes: Configure your CMS or word processor to use straight quotes. 2. Avoid em dashes: Use a hyphen instead of — (em dash). 3. Avoid emoji in cost-sensitive campaigns: If you must use emoji, know it will halve your chars/segment. 4. Test before sending: Use our character counter tool to preview segment count. 5. Use message templates: Pre-validated templates ensure you never accidentally include UCS-2 characters. 6. Review copy-pasted text: Text copied from Word, Google Docs, or websites often contains smart quotes and other UCS-2 triggers.