Encodings
| Options | Standard | ||||
|---|---|---|---|---|---|
| Binarybinary | 2 chars | overhead +700% | checksum no | separate?, symbols?, bits?, order? | Base2, most significant bit first |
| Octaloctal | 8 chars | overhead +172% | checksum no | separate? | Byte values in base 8 |
| Decimaldecimal | 10 chars | overhead +156% | checksum no | codepoints?, separate? | Byte values in base 10; Unicode code points |
| Integerinteger | 10 chars | overhead +141% | checksum no | radix?, digits?, bijective?, width? | Unsigned big-endian integer in a positional numeral |
| Hex (Base16)hex | 16 chars | overhead +100% | checksum no | codepoints?, upper? | RFC 4648 §8; Unicode code points |
| Base32base32 | 32 chars | overhead +60% | checksum no | alphabet?, padding?, canonical? | RFC 4648 §6 and §7, Crockford's Base32, z-base-32 |
| Base45base45 | 45 chars | overhead +48% | checksum no | no options | RFC 9285 |
| Base58base58 | 58 chars | overhead +37% | checksum option | alphabet?, check?, blocks? | Bitcoin Core base58.cpp, Monero common/base58.cpp, Flickr short URLs, XRP Ledger addresses |
| Base64base64 | 64 chars | overhead +33% | checksum no | alphabet?, padding?, canonical? | RFC 4648 §4 and §5 |
| Base85base85 | 85 chars | overhead +25% | checksum no | alphabet?, delimiters? | RFC 1924 alphabet in git's grouping, Adobe Ascii85, ZeroMQ RFC 32/Z85 |
| basE91base91 | 91 chars | overhead +23% | checksum no | no options | Joachim Henke, basE91 reference implementation (base91.sourceforge.net) |
| Base256base256 | 256 chars | overhead +0% | checksum no | alphabet?, symbols?, sample?, multibase? | multiformats base256emoji |
| Bech32bech32 | 32 chars | overhead +62% | checksum yes | prefix, limit?, m? | BIP173, BIP350 (Bech32m) |
| uuencodeuuencode | 64 chars | overhead +38% | checksum no | name?, mode? | POSIX uuencode(1) and uudecode(1) |
| Quoted-Printablequoted-printable | 96 chars | overhead +125% | checksum no | no options | RFC 2045 §6.7 |
| HTML character referenceshtml | 96 chars | overhead text only | checksum no | ascii? | WHATWG HTML, named and numeric character references |
| ITA2 (Baudot-Murray)ita2 | 2 chars | overhead text only | checksum no | alphabet?, order?, numbers?, separate? | ITU-T S.1 (ITA2); CCITT-1 (ITA1); US-TTY |
| Prefix codeprefix | 2 chars | overhead text only | checksum no | codebook?, code?, n?, separate? | Kraft-McMillan inequality; Elias, IEEE Trans. Inf. Theory 21 (1975) |
| Chess squaressquare | 16 chars | overhead text only | checksum no | layout?, pairs?, bitboard? | FIDE Laws of Chess, Appendix C (algebraic notation) |
| FENfen | 34 chars | overhead text only | checksum no | layout?, diagram?, bitboards?, turn?, castling?, enPassant?, halfmove?, fullmove? | PGN Standard §16.1 (Forsyth-Edwards Notation) |
Every row is create(name).info(), computed in your browser from the registry the CLI and the agent tools read. Sort by overhead and watch binary sit at the bottom with +700%. Sort by alphabet and base256 wins with 256 emoji. Among the printable ASCII ones basE91 wins, Quoted-Printable cheats.
Picking one
Need it to survive a URL? base64 with the url alphabet. An email body? base64, which MIME already speaks. A person reading it aloud over a bad phone line? base32 with Crockford's alphabet or z-base-32.
Something money gets sent to? Anything with a checksum. Base58Check catches most typos, bech32 and bech32m catch any four. Plain base58 catches none, and that's why it never travels alone in Bitcoin.
Matching what somebody else wrote? Their spec wins, every time. A QR code wants base45. A PDF stream wants Ascii85. ZeroMQ wants Z85. They look alike and they're not interchangeable.
Squeezing bytes into the fewest printable characters? basE91, at about 23% overhead. Nobody else will read it, so make sure you're the only reader.