Online Encoders and Decoders

Encoding is not encryption

Encoding changes how bytes are represented. It does not protect them. Base64, hex, URL escaping and HTML entities are all reversible by anyone, with no key, and that is the entire point: they exist so data survives a channel that would otherwise mangle it. ROT13 and the Caesar cipher belong here for puzzles and spoiler-hiding, not for secrets. If something needs to be unreadable to whoever intercepts it, you want hashing or encryption from the hash and crypto tools instead.

Picking the right encoding

• Binary in a text field. Base64 is the default: encode or decode text, turn a file into Base64 or back again, with dedicated PDF and Base64 to PDF paths. It costs roughly a third more bytes, because every three bytes become four characters.
• Binary in a URL, filename or address. Standard Base64 uses + and /, which are not URL-safe. Base32 is case-insensitive and survives systems that fold case; Base58 drops the characters people misread, which is why Bitcoin addresses use it.
• Text in a URL. URL encoding has two modes that are not interchangeable. One escapes a whole URL and leaves /, ? and & working; the other escapes a single query value, where those same characters must be escaped or they break the query.
• Text in HTML. HTML entity encoding turns <, > and & into entities so markup is displayed rather than parsed.
• A small asset inline. The data URI generator embeds a file directly in CSS or HTML, trading one network request for a larger, separately uncacheable document.
• A domain with non-ASCII characters. Punycode is the ASCII form DNS actually carries. Email bodies and headers use quoted-printable, which leaves readable text alone and escapes only what it must.

Reading bytes you did not expect

When text arrives as ’ where an apostrophe should be, the bytes are usually fine and the decoding is wrong: UTF-8 read as a single-byte encoding. The UTF-8 byte inspector shows the code points and the byte sequence behind each character, which is the fastest way to tell a mis-decode from genuinely corrupt data. Hex to text, binary to text, the Unicode converter and Unicode escapes cover the same work in the other representations, and when you have a string and no idea what it is, the multi-format encoder runs several at once so you can recognise the one that produces sense.