What alphabet does standard Base32 use?

RFC 4648 Base32 uses the 32 characters A through Z plus the digits 2 through 7. It deliberately omits 0, 1, 8, and 9 to avoid confusion with the letters O, I, and B, making the encoding safer to read aloud or transcribe by hand.

Why is Base32 used instead of Base64?

Base32 is case-insensitive and avoids ambiguous characters, so it survives systems that uppercase text or that humans must read and type. It is common in TOTP two-factor secret keys, DNS records, and file-sharing hashes, trading a larger size for robustness.

What is the = padding for?

Padding with the equals sign pads the encoded output to a multiple of 8 characters, which keeps fixed-size blocks aligned for streaming decoders. It carries no data and decoding ignores it, so you can usually turn it off for shorter output.

How much larger does Base32 make data?

Base32 packs 5 bits into each character, so the encoded form is about 1.6 times the size of the original bytes, or 8 characters for every 5 input bytes. That is bigger than Base64 but the extra robustness is often worth it.

What happens if I paste invalid characters?

The decoder rejects any character outside the Base32 alphabet and shows which character was invalid. Whitespace and trailing padding are stripped automatically, so wrapped or padded strings decode without manual cleanup.

What is the Base32 Encoder and Decoder?

Encode any text to standard RFC 4648 Base32 using the A-Z and 2-7 alphabet, or decode a Base32 string back to text. Optional = padding, whitespace-tolerant decoding, and clear error messages for invalid characters. Runs in your browser. It runs free in your browser on Gera Tools, with nothing uploaded.

Base32 Encoder and Decoder

Name: Base32 Encoder and Decoder
Creator: Gera Tools
License: https://creativecommons.org/licenses/by/4.0/

Get one useful tool a week

Like this tool? Enter your email and we'll send you one genuinely useful Gera tool a week — plus a link to come back to this one. No spam, one-click unsubscribe any time.

Base32 is a binary-to-text encoding that represents data using only the uppercase letters A to Z and the digits 2 to 7. This tool encodes text into that form and decodes it back, handling padding and stray whitespace for you.

How it works

Encoding reads the input bytes as a continuous stream of bits and slices it into 5-bit groups, mapping each group to one alphabet character:

alphabet = ABCDEFGHIJKLMNOPQRSTUVWXYZ234567   (index 0..31)
take 8 bits at a time, emit a character for every 5 bits accumulated
pad the final group with zero bits, then add '=' to reach a multiple of 8

Decoding reverses this: each character becomes its 5-bit index, the bits are re-concatenated, and every full 8 bits becomes one output byte. Leftover bits from padding are discarded.

Worked example and tips

The text Gera Tools (10 bytes) encodes to a Base32 string. Each 5 bytes of input becomes 8 characters of output, so 10 bytes fills exactly 16 characters with no padding needed. A 6-byte input would need 2 padding characters to reach the next multiple of 8.

Because the RFC 4648 Base32 alphabet deliberately excludes 0, 1, 8, and 9, it is the format behind TOTP authenticator-app secrets: you can dictate the key over the phone digit by digit without the listener confusing 0 and O or 1 and I. The same property makes Base32 reliable in DNS TXT records, where some systems quietly uppercase or lowercase text.

Base32 in context: when to choose it

Format	Case sensitivity	Avoids look-alikes	Space overhead	Common use
Base64	Case-sensitive	No (uses `+`, `/`)	~33%	Email, data URIs, JWTs
Base32	Case-insensitive	Yes (A-Z, 2-7)	~60%	TOTP keys, file hashes, DNS
Base32Hex	Case-insensitive	Partial	~60%	Sort-order–preserving IDs
Hex (Base16)	Case-insensitive	N/A (digits + A-F)	100%	Byte dumps, cryptographic hashes

Choose Base32 when:

The output will be read or typed by a human
The system stores or transmits the value as case-insensitive text
You need the encoded form to survive email forwarding, DNS propagation, or legacy uppercasing middleware

If you are storing a value where length matters more than legibility, prefer Base64. If you need a URL-safe compact token, Base62 or URL-safe Base64 are better fits. Everything runs locally in your browser; nothing is uploaded.