What is UCS-2 encoding?
UCS-2 is the 16-bit Unicode encoding SMS falls back to for emoji, Cyrillic, Arabic, or CJK text — 70 characters per message instead of 160, and surrogate-pair emoji count double.
Definition
UCS-2 is the 16-bit encoding an SMS switches to when its text cannot be expressed in the GSM-7 alphabet — any Cyrillic, Arabic, Hebrew, CJK, Indic text, or a single emoji triggers it. At two bytes per character, the 140-byte SMS payload fits 70 characters in one message, or 67 per part of a concatenated message once the reassembly header is deducted.
Two technical wrinkles matter in practice. First, billing counts UTF-16 code units, not visible characters: emoji outside Unicode’s Basic Multilingual Plane are encoded as surrogate pairs and cost two of your 70, and multi-codepoint emoji (skin tones, flags, families) cost more still. Second, the switch is all-or-nothing per message — 199 GSM-7 characters plus one emoji is a three-segment UCS-2 message, not two GSM-7 segments plus a little. For languages that need Unicode there is no way around UCS-2; the optimisation is keeping messages short and not paying the Unicode price by accident on messages that didn’t need it.
Why it matters
UCS-2 is either a necessity (non-Latin scripts) or an accident (one pasted smart quote) — and the accident version silently multiplies campaign costs. Knowing which of the two you are paying for, per template, is basic messaging hygiene.
See it in action
See which characters force UCS-2 →