What is Base62 good for?

Base62 uses only the 62 characters that are letters and digits, so the output never contains symbols that must be escaped in a URL, filename, or HTML attribute. That makes it a popular choice for short link slugs and compact public IDs derived from large database keys.

How is Base62 different from Base64?

Base64 has 64 symbols including plus and slash, which are not URL-safe, and it pads with equals. Base62 drops those three problem characters, leaving a clean alphanumeric set. The trade-off is that Base62 cannot use simple fixed-width bit groups, so it relies on big-integer base conversion instead.

When should I use integer mode versus bytes mode?

Use integer mode when you are shortening a numeric identifier such as an auto-increment primary key, because it encodes the number directly with no padding. Use bytes mode to round-trip arbitrary text or binary data, where each leading zero byte is preserved as a leading zero character.

Does Base62 preserve leading zeros?

In bytes mode, yes: each leading zero byte is encoded as a leading '0' character and restored on decode. In integer mode there are no leading zeros to preserve, since a number like 7 simply encodes to a single character.

Is my data uploaded anywhere?

No. Encoding and decoding run entirely in your browser using JavaScript and BigInt, so your IDs and data stay on your device.

Base62 Encoder/Decoder

Base62 encodes data using the 62 characters 0-9, A-Z, and a-z — every alphanumeric character and nothing else. With no symbols to escape, the output drops cleanly into URLs, filenames, and HTML attributes, which is why it is a favourite for short link slugs and compact IDs. This tool encodes integers or UTF-8 bytes to Base62 and decodes them back, all in your browser.

How it works

Base62 is positional base conversion against the alphabet 0-9A-Za-z:

In integer mode the input number is divided by 62 repeatedly; each remainder picks one character, read most-significant first. JavaScript BigInt is used so arbitrarily large IDs encode exactly.
In bytes mode the input bytes are treated as one big-endian integer and converted the same way, with each leading zero byte emitted as a leading 0 so the value round-trips.
Decoding accumulates each character’s value back into the integer, then either prints the number (integer mode) or rebuilds the bytes (bytes mode).

Edge cases are handled explicitly: the number 0 encodes to a single 0, empty input yields empty output, and any character outside the alphabet is rejected by name.

Tips and example

The integer 1000000 encodes to 4C92 in Base62 — four characters instead of seven digits, which is the whole point when building short URLs. Decoding 4C92 in integer mode returns 1000000. If you need a checksum or a human-transcription-proof alphabet, prefer Base58 instead; Base62 optimises for the shortest possible URL-safe string rather than for resisting typos.