How to Encode Base64

Steps

1. Start with the exact text you want represented as bytes.
2. Encode those characters as UTF-8. Skipping this and using a Latin-1-only function breaks å, ä, ö and emoji.
3. Apply the Base64 alphabet to those bytes.
4. Copy the ASCII result. Remember: anyone can decode it.

On the MultiConvers Base64 encoder, paste the text and choose Encode. The UTF-8 step is already done.

A worked example

“Hello” is five UTF-8 bytes: 72, 101, 108, 108, 111 (hex 48 65 6C 6C 6F). Base64 takes those bytes three at a time. The first three become the characters SGVs. The last two bytes are not a full group of three, so the encoder pads with one =. The whole string is SGVsbG8=.

“Göteborg” is longer in UTF-8 than in letters, because ö is two bytes. That is why its Base64, R8O2dGVib3Jn, is not a one-byte-per letter encoding. Skip the UTF-8 step and those characters corrupt.

The limit

This page encodes text. A file’s raw bytes use the same alphabet only after something else has read those bytes; pasting a filename does not encode the file. The result is reversible ASCII, not a secret. Anyone can decode it without a key. Stripping the trailing = can make a strict decoder reject the string, so keep the padding.

Examples

Base64 is still not encryption — see what Base64 is. If you need a value safe inside a query string, you may also need percent-encoding.