URL Encoder & Decoder

How to encode and decode a URL value

1. 1

   Choose a mode

   URL component for a query value or path segment. Form value for one value of form data.

2. 2

   Add one value

   Type, paste or import it. The whole input is processed as one value, exactly as written.

3. 3

   Press Encode or Decode

   Check the output, then copy it, download it, or use it as the next input.

The two modes

• URL component. Works like JavaScript’s encodeURIComponent and decodeURIComponent. A space becomes %20 and a plus becomes %2B. When decoding, + stays a plus sign. Example: a b+c becomes a%20b%2Bc.
• Form value. Follows application/x-www-form-urlencoded, as HTML forms send it, using the browser’s URLSearchParams. A space becomes + and a plus becomes %2B. When decoding, + becomes a space. Example: a b+c becomes a+b%2Bc. It handles one value, not a whole query string, so a literal & is refused when decoding.
• Why there is no “encode a whole URL” option. Encoding a complete URL as a component also encodes its structure: https://example.com/a?b=c becomes https%3A%2F%2Fexample.com%2Fa%3Fb%3Dc, which no longer works as a link. Encode each value on its own, then join the parts.
• Reserved characters. & = / ? # and : are always encoded (as %26, %3D, %2F, %3F, %23, %3A), so a value cannot break out of its place in a URL. Decoding turns them back.

What to expect

• Unicode. Accents, emoji and other scripts are turned into UTF-8 bytes, then each byte is written as %XX. Decoding reads the bytes back as UTF-8.
• One pass per press. Existing % signs are encoded again (%20 becomes %2520). If a value was encoded twice, press Decode a second time; the tool never repeats decoding by itself.
• Malformed input is refused. A % without two hex digits, escapes that are not valid UTF-8, and lone surrogate characters are reported with their position. Nothing is repaired or replaced silently.
• Nothing is changed behind your back. Spaces and line breaks at either end are kept, lines are not split, and nothing is converted while you type. Empty input gives an empty output and says so. If an action fails, the earlier output stays and is labelled out of date.

Limits

largest input, in UTF-8

1 MB

Larger input is refused with a message rather than cut. The work runs in a background thread (a Web Worker) so the page stays responsive. Text boxes store line breaks as LF, so a line break you type is encoded as %0A.

Frequently asked questions