Base58 Encoder and Decoder Online

How to encode to Base58

1. Type or paste text into Input · Text. The Base58 string appears in Output · Base58 as you type.
2. Choose the Alphabet under the output: Bitcoin (the default, also used by IPFS and Solana), Flickr or Ripple.
3. To encode raw bytes instead of text, set From to Hex or Base64 and paste the bytes in that form. A public key or hash is almost always bytes, not text.
4. Click Base58 → Text in the direction toggle to decode, or open the Base58 decoder, which shows the bytes as hex.

How Base58 encoding works

Base58 does not cut the input into fixed bit groups the way Base64 does. It reads all the bytes as one big-endian number and writes that number in base 58, then adds one 1 for every leading zero byte. That is why a single changed byte can change the whole string, and why the length depends on the value and not only on the byte count.

The cost of treating the data as one number grows quickly with size, so this tool caps Base58 input at 64 KB. That is far more than any key, hash or address needs.

The three Base58 alphabets

All three use the same 58 characters (digits and letters without 0, O, I, l), in a different order:

Flickr swaps upper and lower case, which Flickr used for short photo URLs. Ripple’s alphabet starts with r, so XRP Ledger addresses, which have a zero version byte, start with r the way Bitcoin addresses start with 1. Decoding with the wrong alphabet usually works without an error and gives different bytes, so check the alphabet first when a value does not match.

Leading zeros and the 1 prefix

Because Base58 treats data as a number, zero bytes at the front would vanish. The format keeps them by writing one 1 per zero byte:

The last row is a real address: thirty-two 1s is the Solana System Program, whose public key is 32 zero bytes. Leading zeros are where hand-written Base58 code most often goes wrong, and an encoder that drops them produces strings that decode to fewer bytes.

Where Base58 is used

• Bitcoin addresses and WIF keys. Legacy addresses (1…, 3…) and WIF private keys (5…, K…, L…) are Base58Check: Base58 with a version byte and a checksum. Bech32 addresses (bc1…) are a different encoding.
• IPFS CIDv0. The Qm… identifiers are plain Base58 of a 34-byte multihash: 12 (SHA-256), 20 (32 bytes long) and the digest. The 12 20 header is what makes every CIDv0 start with Qm.
• Solana. Addresses and public keys are Base58 of 32 bytes (32 to 44 characters); transaction signatures are Base58 of 64 bytes (up to 88 characters).

Base58 in Python

Python has no Base58 in the standard library. The algorithm is short enough to write out, and this version handles leading zeros:

ALPHABET = "123456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnopqrstuvwxyz"

def b58encode(data: bytes) -> str:
    n = int.from_bytes(data, "big")
    out = ""
    while n:
        n, r = divmod(n, 58)
        out = ALPHABET[r] + out
    zeros = len(data) - len(data.lstrip(b"\0"))
    return "1" * zeros + out

b58encode(b"Hello")           # '9Ajdvzr'
b58encode(bytes(32))          # '11111111111111111111111111111111'

For production code, a maintained package such as base58 on PyPI also covers Base58Check. General Python encoding patterns are on the Python page.

FAQ