Whether a "100 characters or fewer" limit means character count or byte count changes how much non-English text you can actually fit.
What character count means
This counts each visible character one by one. "Hi" is 2 characters, and so is "์๋ " โ the writing system doesn't matter.
Why byte count differs
In UTF-8, English letters and digits take 1 byte, but a Korean syllable block takes 3 bytes. Emoji often take 4 bytes depending on the character. So text mixing in Korean occupies far more bytes than English text with the same character count.
When byte count actually matters
Sending SMS (billed and split by byte count), designing a database column size (in a database where VARCHAR is byte-based), and character limits on social platforms like X (which vary between character-based and byte-based counting) all need the byte count checked, not just the character count.