Base64 Decoder

Decoding Base64 sounds like it should always produce readable text, but that's only true if the original content actually was text. Base64 is just as often used to represent binary data — an image, a font file, a compressed blob — and a decoder that blindly treats every result as text will show you a screen full of garbled characters instead of something useful.

This tool inspects the decoded bytes before deciding how to show them. If the byte pattern matches a known image format's file signature — the same "magic bytes" browsers and operating systems use to identify file types — it renders the result as an actual image you can view and download, instead of a string of nonsense characters. If the bytes look like readable text, it decodes them properly; if that text also happens to parse as valid JSON, it flags that too, since a Base64-encoded JWT payload or API response is an extremely common thing to decode.

Text decoding itself is UTF-8 correct, using the browser's TextDecoder rather than the plain atob() function many simpler decoders rely on. atob() only understands Latin1 byte values, so decoding anything encoded from Unicode text — emoji, accented characters, non-Latin scripts — through a naive decoder produces corrupted output. This tool decodes the underlying bytes properly as UTF-8, matching how the paired Base64 Encoder on this site actually encodes text in the first place.

A URL-safe toggle handles Base64 that uses the - and _ alphabet (common in JWTs and URL-embedded tokens) instead of the standard + and / — and the tool detects URL-safe characters automatically in most cases, so you often don't need to touch the toggle at all. Everything happens locally: your Base64 string is decoded entirely in your browser, and any binary data never has to leave your device to be downloaded.