Multibase Encoding Explained

Problem

You see strings prefixed with single letters (like b, f, B) and need to understand the multibase convention that self-identifies the encoding.

Prerequisites

• Familiarity with Base64, Base32, and hex
• Understanding of self-describing formats

Steps

1. 1. Learn the prefix table

   Multibase prepends a single character to indicate the encoding: B for Base64, b for Base64url, f for hexadecimal, F for uppercase hex, z for Base58 Bitcoin, and K for Base36. This lets a consumer detect the encoding without external metadata.

   Input

   f4d2

   Output

   hex encoding of 0x4d2

2. 2. Encode with the correct prefix

   Choose the encoding that fits your needs, encode the data, and prepend the corresponding prefix. For binary data in URLs, use b (Base64url). For compact human-readable codes, use z (Base58).

   Input

   binary data

   Output

   bAAEC (Base64url with prefix)

3. 3. Decode by inspecting the prefix

   Read the first character to determine the encoding, strip it, and decode the rest using the identified base. A multibase decoder handles all supported prefixes automatically.

   Input

   bAAEC

   Output

   decode as Base64url → bytes [0,1,2]

Expected result

A self-identifying encoded string that any multibase-aware consumer can decode without prior knowledge of the encoding.

Common errors

Forgetting to strip the prefix before decoding

The first character is the encoding marker, not part of the data. Remove it before passing to the base-specific decoder.

Using the wrong prefix for the encoding

Refer to the multibase table. Using B (Base64) when you meant b (Base64url) will produce different output.

Troubleshooting

• If decoding fails, check the prefix against the multibase table to confirm it matches the actual encoding.
• Multibase is used in IPFS CIDs; the prefix is part of a larger self-describing structure.

Security limits

• Multibase is a transport encoding. It provides no confidentiality or integrity guarantees.