base58

A fast, zero-dependency Base58 encoding library for Go with support for both standard and chunked encoding.

• Standard Base58 — variable-length encoding commonly used for Bitcoin addresses
• Chunked Base58 — fixed-block encoding that splits input into 8-byte chunks, providing O(n) performance instead of O(n²) for large payloads
• Buffer reuse — EncodeTo allows encoding into a caller-provided buffer with zero allocations
• Custom alphabets — includes Bitcoin and Flickr alphabets, or create your own with NewEncoding

go get github.com/KarpelesLab/base58

// Encode
encoded := base58.Bitcoin.Encode(data)

// Decode
decoded, err := base58.Bitcoin.Decode(encoded)

// Encode with buffer reuse (zero allocations)
buf := make([]byte, base58.EncodedMaxLen(len(data)))
result := base58.Bitcoin.EncodeTo(buf[:0], data)

Chunked encoding processes data in independent 8-byte blocks, making it significantly faster for larger payloads at the cost of a slightly longer output.

// Encode
encoded := base58.Bitcoin.EncodeChunked(data)

// Decode
decoded, err := base58.Bitcoin.DecodeChunked(encoded)

enc := base58.NewEncoding("123456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnopqrstuvwxyz")
encoded := enc.Encode(data)

Standard base58 is O(n²) due to full-buffer carry propagation. Chunked encoding is O(n) since each 8-byte block is encoded independently. Benchmarks on an Intel i9-14900K:

Run benchmarks yourself:

go test -bench=. -benchmem