Base58 Encoder and Decoder

The Bitcoin Base58 alphabet

123456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnopqrstuvwxyz

Notice what is missing: 0, O, I, l. The first character in the alphabet is 1, which represents the integer 0.

How it actually works

Unlike Base64 which chunks bits in fixed groups, Base58 treats the whole byte string as a single big-endian integer N and expresses it in base 58:

result = []
while N > 0:
    result.prepend(alphabet[N mod 58])
    N = N / 58
for each leading zero byte in input:
    result.prepend(alphabet[0])  # i.e. '1'

This is why Base58 encoding is non-trivial to stream, you need the whole number before you can emit any character.

Base58 vs Base58Check

A typical legacy P2PKH Base58Check payload is 25 bytes: 1 version byte, 20 hash bytes and 4 checksum bytes. The checksum is the first 4 bytes of the double-SHA-256 digest of the version and hash. Encoding that complete payload as Base58 produces the familiar address string.

This tool converts plain Base58 only. Use a dedicated Base58Check validator when you need to verify an address version and checksum.

Alphabet variants

Why not just use Base64?

Base64 uses 4 encoded characters for each 3 input bytes before padding, a fixed and efficient expansion. Base58 has slightly variable overhead but avoids several ambiguous characters, which helps when someone copies an identifier by hand or reads it aloud.