What is Base64 Encoding? A Simple Explanation for Everyone

I was debugging an email delivery issue last year when I noticed a block of gibberish in the raw message source: SGVsbG8gV29ybGQ=. My first instinct was that something had gone wrong with the encoding. Turns out, nothing was broken — that was Base64 doing exactly what it's supposed to do. If you've ever bumped into these strange-looking strings in URLs, API headers, or JSON payloads and wondered what on earth they are, this article should clear things up. You can also try encoding or decoding strings yourself with our Base64 Encoder / Decoder.

The Problem Base64 Was Built to Solve

Computers handle everything as binary — sequences of 0s and 1s. Text, images, audio files — all binary under the hood. The catch? Many communication channels, particularly older ones like SMTP email, were designed strictly for plain ASCII text. Raw binary data passing through these channels would get mangled, since certain byte values get misinterpreted as control characters.

Base64 sidesteps this problem entirely. It takes any binary data and converts it into a string made up of just 64 safe characters: A-Z, a-z, 0-9, +, /, and = for padding. Every system on the planet can handle these characters without corruption.

The Base64 Alphabet

Here's the full character set. Each character represents a value from 0 to 63:

Some implementations use different characters for 62 and 63. URL-safe Base64, for example, swaps + with - and / with _ so the encoded string works inside URLs without breaking anything.

How the Encoding Process Works, Step by Step

The algorithm is straightforward once you see it done manually. Let's encode the word "Hi":

1. Convert each character to its ASCII binary representation. "H" = 72 = 01001000, "i" = 105 = 01101001
2. Concatenate all bytes into one binary string. 01001000 01101001 → 0100100001101001
3. Pad to a multiple of 6 bits. We have 16 bits. Next multiple of 6 after 16 is 18, so add two zero bits: 010010000110100100
4. Split into 6-bit groups. 010010 | 000110 | 100100
5. Map each group to the Base64 table. 18→S, 6→G, 36→k
6. Add padding. Original input was 2 bytes (not divisible by 3), so we add one = pad.

Final result: SGk=

Common Real-World Uses

Base64 pops up in more places than you'd expect. Here are the situations where developers reach for it most often:

Base64 Is NOT Encryption — Seriously

I can't stress this enough, because I've seen it in production codebases: people using Base64 to "hide" API keys, passwords, or personal data. Base64 provides zero security. There's no key, no algorithm complexity, nothing. Any developer (or anyone with Google, frankly) can decode a Base64 string in under five seconds. If you need to protect data, use actual encryption — AES-256, for example — and store encryption keys securely.

The Size Cost

Every encoding comes with trade-offs, and Base64's is size. Because you're converting 3 bytes of raw data into 4 Base64 characters, the encoded output is about 33% larger than the original.

For a 2 KB favicon, that overhead is negligible. For a 5 MB photo, you're adding 1.65 MB of bloat — at that point, you're better off transferring the file as binary directly.

When You Should (and Shouldn't) Use It

Related Developer Tools

Working with encoded data? These might come in handy alongside the Base64 tool:

• URL Encoder / Decoder — percent-encoding for URLs
• Hash Generator — SHA-256, MD5, and other one-way hashes
• HTML Entity Encoder — escape special characters for HTML
• JSON Formatter — pretty-print and validate JSON payloads