Decoded output is garbled

check whether the original was encoded with `btoa()` (Latin-1 only) or `Buffer.from()` (UTF-8). In the browser, encode UTF-8 safely with `btoa(unescape(encodeURIComponent(text)))`. In Node, use `Buffer.from(text, 'utf8').toString('base64')`.

Output has unexpected characters

you're probably dealing with Base64URL, not standard Base64. Replace `-` with `+` and `_` with `/`, pad to a multiple of 4 with `=`, then decode again.

`btoa()` throws on my string

it can't handle characters outside Latin-1. Emoji, CJK, anything with code points above 255 will fail. Use the `encodeURIComponent` workaround above or switch to Node's `Buffer`.

String has line breaks in it

that's valid. MIME Base64 wraps at 76 characters. Most decoders handle it fine, but strip whitespace first if yours doesn't.

Decoded file is corrupt

almost always trailing whitespace or a stray newline that got copied with the string. Trim and retry.

browsers support large data URIs but inlining a 500 KB image as Base64 in your CSS is a bad time. Keep data URIs under 5–10 KB. Anything bigger should be its own file request.

All Tools

toolzy.dev / Encoders & Decoders / Base64-Encoder

T

Base64 Encoder/Decoder

Encode text to Base64 or decode Base64 strings.

Input

Output