URL Encoder/Decoder

Component encoding (encodeURIComponent) escapes everything except unreserved characters — A-Z, a-z, 0-9, and - _ . ~ — plus a handful of others it leaves alone (! * ' ( )) — use it for a single query parameter or path segment. Full URL encoding (encodeURI) additionally leaves reserved, URL-structural characters untouched (: / ? # [ ] @ & = + $ , ;), since encoding those would break the URL itself — use it for an entire URL.

The input isn't valid percent-encoded text — for example it contains a lone % not followed by two hex digits.

It's the percent-encoded form of a space character — %20 is hex 20, the ASCII code for space. You'll see it anywhere a URL needs to represent a literal space.

That's a different, older encoding scheme — application/x-www-form-urlencoded, used by HTML forms in query strings — where a literal + means space. This tool uses standard URI encoding (encodeURI/encodeURIComponent), which always produces %20 for a space, never +.

Non-ASCII characters, emoji included, get converted to UTF-8 bytes first, then each byte is percent-encoded separately. A 4-byte emoji becomes four %XX groups — 12 characters from one character.

Unreserved characters (A-Z, a-z, 0-9, - _ . ~) are always safe and never need encoding. Reserved characters (: / ? # [ ] @ ! $ & ' ( ) * + , ; =) have special meaning in a URL's structure, so they're left alone only in Full URL mode and encoded in Component mode.

Whenever it might contain spaces, symbols, or non-English text — for example a search query, a name, or a redirect target being passed as a query parameter. Encoding it first keeps the URL valid and prevents it from breaking the parameters around it.

Yes — paste the percent-encoded value into Decode mode with the matching scope (Component for a single parameter value, Full URL for a complete URL) to get the original text back.

%40 is the percent-encoded form of the @ symbol — commonly seen when an email address is passed as a query parameter value, since @ has special meaning in a URL's structure and needs escaping outside that context.

%27 is a percent-encoded apostrophe ('). It shows up often in encoded names, contractions, or quoted strings passed through a URL.

%23 is the percent-encoded form of #. Since # marks the start of a URL fragment (like a page anchor), a literal # inside a value has to be encoded as %23 so it isn't mistaken for the start of a fragment.

%5D is a right square bracket ], and %5B is a left square bracket [. Both are reserved characters, encoded when they appear inside a value rather than as part of the URL's own structure.

%7D is a right curly brace }, and its counterpart %7B is a left curly brace {. Neither has a reserved meaning in a URL, but both fall outside the unreserved character set, so encoding a value containing them escapes both.

%0A is a line feed (newline) character — the encoded form of a line break inside a text value passed through a URL, such as a multi-line form field submitted as a query parameter.

There's no decrypting involved — percent-encoding isn't encryption, it's a reversible way to represent special characters safely, with no secret key required. Switch this tool to Decode mode and it converts the percent-encoded text straight back to plain text.

No — they're different schemes for different purposes. Percent-encoding (what this tool does) escapes specific characters so text is safe inside a URL. Base64 re-encodes arbitrary binary data into a larger set of ASCII characters, commonly used for things like embedding images or auth tokens, not specifically for URL safety.

No — HTML encoding (using entities like &amp; or &lt;) escapes characters that would otherwise break HTML markup, like < and &. URL encoding (percent-encoding) escapes characters that would break a URL's structure. They solve different problems and use completely different formats.

Yes — running Encode again on already-encoded text will also encode its % characters, turning something like %20 into %2520. If you're not sure whether text is already encoded, decode it first to check, or compare the input against the expected plain-text value before encoding.