MultiHash

Multihash is a self-describing encoding for cryptographic hash digests used in distributed and content-addressed systems. It provides a portable way to identify both the hash function that produced a digest and the length of that digest, enabling software to verify data without relying on external metadata. Multihash is part of the broader multiformats project, which aims to standardize data formats for interoperable communication.

The multihash format consists of three parts: a code identifying the hash function, a length field, and the digest itself. Both the code and the length are encoded as unsigned varints, a compact variable-length integer encoding. The code specifies the hash function (for example, sha2-256, sha3-512, or blake2b-256). The length indicates how many bytes are in the following digest, and the digest is the raw output produced by applying the corresponding hash function to the input data. This structure makes the multihash self-describing: given the multihash, software can determine how to verify the data and which algorithm to use.

Multihash is widely used in IPFS and related technologies, where content addressing is central. It supports upgrading to stronger hash functions over time without breaking existing data, since the multihash itself encodes the algorithm and length. Representations of multihashes can be encoded in various base encodings, such as base58btc or base32, through the multibase system. A common example for SHA-256 is a multihash beginning with the code 0x12, followed by the length 0x20, and then the 32-byte digest.