Base36 Encoder and Decoder

What is Base36?

Base36 writes data with the ten digits and the 26 letters of the Latin alphabet, 0-9 followed by a-z. It is the largest alphabet you can build from digits and letters while ignoring letter case, so Base36 strings survive systems that change or ignore case. Each character carries about 5.17 bits, which makes the output roughly 1.55 times the size of the input, a little shorter than Base32.

Because 36 is not a power of two, the bytes cannot simply be cut into bit groups. Instead the tool reads the whole input as one large unsigned number and converts it to base 36, the same method Base58 uses. A number would lose its leading zero bytes, so each one is written as a leading 0 and restored when decoding. This arithmetic gets slow for large inputs, so the tool accepts up to about 20 KB.

Where Base36 is used

• Short IDs. URL shorteners, order numbers and ticket codes often show a numeric database ID in base 36. Reddit, for example, uses base 36 for its post and comment IDs.
• Case-insensitive places. DNS names and some file systems ignore letter case. IPFS uses a Base36 form of some identifiers in subdomain gateway addresses for this reason.
• Numbers in code. Base 36 is the highest radix that most programming languages support natively for integer conversion.

If letter case is preserved and you want the shortest alphanumeric string, Base62 is more compact.

Base36 numbers in code

This page encodes bytes. To convert an integer, use the built-in functions of your language or the number base converter:

// JavaScript
(1000000).toString(36)        // 'lfls'
parseInt('lfls', 36)          // 1000000
123456789012345678901234567890n.toString(36)   // BigInt for large values

# Python
int('lfls', 36)               # 1000000
# there is no built-in for the reverse direction

parseInt loses precision above Number.MAX_SAFE_INTEGER; for larger values, accumulate the digits into a BigInt yourself.