Developer guides
Base64 Encoding Explained: Examples, UTF-8, Base64URL
Updated 25 September 2026 · 4 min read
Base64 turns any bytes — an image, a PDF, text in any language — into plain letters and digits that survive email, JSON, URLs and HTML. It's everywhere: data URLs, email attachments, API keys and the parts of a JWT.
How it works
Base64 takes 3 bytes (24 bits) at a time and splits them into four 6-bit groups. Each 6-bit value (0–63) maps to one of 64 characters: A–Z, a–z, 0–9, + and /.
| Text | Base64 |
|---|---|
| Hi! | SGkh |
| Namaste | TmFtYXN0ZQ== |
| नमस्ते (UTF-8) | 4KSo4KSu4KS44KWN4KSk4KWH |
“Hi!” is exactly 3 bytes, so it becomes 4 characters with no padding. “Namaste” is 7 bytes, so it's padded with ==.
Indian languages and emoji: use UTF-8
नमस्ते is 6 characters but 18 bytes in UTF-8. JavaScript's btoa() only handles single-byte characters and throws an error on Hindi or emoji — encode to UTF-8 first (new TextEncoder().encode(text)). The Base64 tool does this for you.
Base64URL
Standard Base64 uses +, / and =, which have special meanings in URLs. Base64URL swaps them for - and _ and usually drops the padding. JWTs use it — see how to decode a JWT.
When to use it (and when not to)
- ✅ Embedding small images or fonts in CSS/HTML, sending files inside JSON, basic-auth headers.
- ❌ Hiding passwords or secrets — anyone can decode it.
- ❌ Large files — Base64 makes data about 33% bigger (4 characters for every 3 bytes).
Frequently asked questions
Is Base64 encryption?
No. Base64 is an encoding anyone can reverse instantly. It makes binary data safe to send as text; it does not hide it.
Why does Base64 end with = or ==?
Base64 works in groups of 3 bytes. When the input isn't a multiple of 3, one or two = characters pad the output to a multiple of 4 characters.