UTF-8 uses a variable number of bytes. ASCII text commonly uses one byte per code point, while accented letters and emoji use more. WordCount measures the exact TextEncoder output instead of multiplying a character count by an assumed width.
UTF-8 is variable-width
A byte limit cannot be inferred from visible length. The combining example e plus an acute mark is one grapheme, two code points and three UTF-8 bytes. The family emoji is one grapheme but twenty-five bytes. Encode the exact final text, including line endings, before comparing with a byte quota.
Storage and transport may transform text
A form library can normalize line endings, a database can apply another encoding, and a protocol can add quoting or escape sequences. This counter measures only the UTF-8 encoding of the text currently in the field. It does not include JSON quotes, URL escaping, compression, headers or database overhead.
Verify the receiving system
Some products advertise a character limit but enforce UTF-16 units or bytes. Others count normalized text after submission. Use a small boundary fixture containing ASCII, an accented character and an emoji, then confirm what the real endpoint accepts. WordCount can expose candidate measures but cannot certify an external implementation.