Open the converter, choose Encode, Standard Base64 and Padded, then enter exactly Hello, мир 🌍. Do not add a line ending. The source contains 13 UTF-16 code units, but it becomes 18 UTF-8 bytes because Cyrillic letters and the emoji require multiple bytes. Base64Lens shows the byte preview before producing the canonical encoded text.

Exact UTF-8 bytes

The expected hexadecimal evidence is 48 65 6C 6C 6F 2C 20 D0 BC D0 B8 D1 80 20 F0 9F 8C 8D. ASCII occupies one byte per character, each Cyrillic letter here occupies two, and the globe emoji occupies four. This distinction explains why visible character count cannot predict Base64 length for multilingual text.

Canonical output has no equals signs

The exact standard Base64 output is SGVsbG8sINC80LjRgCDwn4yN. Eighteen bytes form six complete three-byte groups, so padded and unpadded output happen to match. This does not mean padding was ignored: a different byte count would make the chosen policy visible at the end.

Decode back without normalization

Switch to Decode with the same alphabet and padded policy, paste the output and confirm that the exact original text returns. Compare spaces, comma, Cyrillic spelling and emoji. Base64Lens does not normalize Unicode or change line endings, so any extra character would correctly change the byte evidence and encoded result.