URL Encoder, Decoder and Parser

Percent-encoding replaces characters that are not safe in a URL with a percent sign and the hexadecimal value of each UTF-8 byte, so ë becomes %C3%AB. Which characters are left alone depends on the job, and the mode selector covers the three you meet in practice. Component mode is JavaScript's encodeURIComponent: it escapes everything except letters, digits and - _ . ! ~ * ' ( ), which is right for a single query value, a path segment or a redirect target.

Whole-URL mode is encodeURI. It keeps ; , / ? : @ & = + $ # so a complete URL stays a URL, which means it will not protect an & or = that sits inside a value. Form mode produces application/x-www-form-urlencoded as the WHATWG URL standard defines it: spaces become +, and only letters, digits and * - . _ remain unescaped. That is what an HTML form or URLSearchParams sends, and it is why + in a query string means a space but in a path is a literal plus.

Decoding fails on a percent sign that is not followed by two hex digits, such as the 100% in a sentence, or on byte sequences that are not valid UTF-8, typically text encoded in Latin-1 where é is %E9 instead of %C3%A9. Instead of a vague URIError the tool points at the exact position and sequence, explains the likely cause, and shows a lenient decode that leaves the broken parts as they were.

Double encoding happens when an already-encoded value is encoded again: %20 turns into %2520. It is a frequent cause of links that show literal %20 in the page or APIs that reject a request. The tool flags text that looks encoded more than once, lists each decoding step, and warns when you are about to encode text that already contains percent-escapes.

Paste a full URL to split it with the browser's own URL parser into protocol, user, host, port, path, query parameters and fragment. Query parameters appear as editable rows; change, add or remove them and the rebuilt URL updates. Internationalised domain names are shown in readable Unicode and in their punycode form (xn--mnchen-3ya.de for münchen.de), which is what DNS and certificates use. Seeing both helps spot look-alike domains, although the tool does not judge whether a name is deceptive.

Rebuilt URLs write parameters in form encoding, so a space becomes +, and duplicate names such as tag=a&tag=b are kept in order. A URL with no scheme is reported as such rather than guessed at.