Base64 Encoder & Decoder

Base64 works on bytes, not characters:

1. Encode: your text is first converted to UTF-8 bytes with TextEncoder (this is what keeps non-English text intact), then every 3 bytes are regrouped into 4 six-bit values, each mapped to a character in A–Z a–z 0–9 + /. If the byte count isn't a multiple of 3, = padding is added.
2. Decode: the process reverses — the 64-character alphabet maps back to bytes, padding is stripped, and TextDecoder rebuilds the original UTF-8 text.

A common bug in naive tools is calling btoa() directly on Unicode text, which corrupts anything outside Latin-1. This tool encodes UTF-8 bytes first, so it doesn't.

Encode the Hindi greeting नमस्ते: you get 4K+4K+4K+4K+4K+-style output (the exact string is 4K+4K+4K+4K+4K+ for a shorter sample — try it live). Decode it back and you get नमस्ते exactly, because the tool round-trips through UTF-8 rather than Latin-1.

Encoded output is safe to paste into JSON, URLs (with care — + and / have URL meanings), headers or code. Decoded output is the original text. If decoding fails, you'll see a clear error — usually the input contains characters outside the Base64 alphabet or has broken padding.

Remember: Base64 is an encoding, not encryption. Anyone can decode it, so never use it to hide secrets.