How to use this tool
- Paste a value, not an entire URL unless you intend to encode all of it.
- Choose encode or decode and the component/form mode.
- Process the text and inspect literal plus and percent signs.
How the operation works
Encoding represents UTF-8 bytes using percent escapes. Component mode writes spaces as %20; form mode uses +. In form decoding, + means space and %2B means a literal plus. Invalid escapes and invalid UTF-8 cause an error rather than replacement characters. Decoding is deliberately one pass.
Encode a component, not an entire URL
A URL contains structural separators such as colon, slash, question mark and ampersand. This tool encodes a component value so those characters cannot accidentally become separators inside that value. Encoding an entire address changes its separators too; the result is suitable as a nested value, not automatically a directly navigable address.
For example, encode a search term before placing it into a query parameter. To inspect a complete address and its repeated query parameters, use URL Parser. To build a campaign link from distinct fields, UTM Builder is more convenient than manually joining encoded strings.
Spaces, plus signs and Unicode examples
Component mode: green tea + mint → green%20tea%20%2B%20mint Form mode: green tea + mint → green+tea+%2B+mint UTF-8 component: é → %C3%A9
In component mode, plus is a literal plus when decoding. In form mode, plus represents a space; an actual plus must be encoded as %2B. Select the mode used by the source system. Guessing the wrong mode can alter names, search terms or other values without producing a syntax error.
Percent escapes represent UTF-8 bytes. One visible character can require several byte escapes. This differs from Unicode code-point notation and Base64 encoding. The encoder also escapes punctuation such as apostrophe and parentheses for component use.
Decode one layer and inspect it
The value %2520 decodes once to %20, not directly to a space. Repeated decoding can change the meaning of data, so this tool performs one explicit pass. Copy the result and run another pass only when you know the source contains another encoding layer.
Incomplete escapes such as % or %2, non-hexadecimal escapes and invalid UTF-8 byte sequences are rejected. Encoding text with an unpaired Unicode surrogate also produces a useful error. It does not silently replace malformed text with a different character.
Limits and handling private values
The input limit is 1 MiB and empty input is rejected. This is a text conversion tool: it does not open a URL, check an HTTP response, confirm a destination or clean tracking parameters. A decoded address can still be unsafe to visit. Read the output before using it in another application.
Input is processed locally and is not placed in the browser address bar. Do not include production secrets in shared examples. Copy and Download are explicit exports; Reset clears the interface. Use HTML entities when the issue is an ampersand reference in markup rather than percent encoding.
A worked example
green tea + mint
Component output: green%20tea%20%2B%20mint
Encoding a whole URL as one component also escapes its structural slashes, colon and question mark. Build full URLs with URL-aware tools instead of concatenating unescaped values.
Processing and privacy
URL Encoder / Decoder processes the text locally; the input and output are not sent to an encoding service. Input stays in browser memory until you reset or leave the page. Copy and download run only when you choose them. No tool input is sent to a processing service. Website pages and scripts still make ordinary hosting requests; the privacy policy separates those from tool processing.
Frequently asked questions
Why is %2520 still %20 after decoding?
It was encoded twice. One decoding pass removes one layer; decide whether another is actually intended.
Should I encode an entire URL?
No—encode component values only. Structural separators must stay unescaped for the address to work.
Why decode only one layer?
Repeated decoding can change meaning. One explicit pass keeps control; decode again only if another layer is intended.
What about invalid escapes?
They are rejected with an error rather than replaced with replacement characters.
Is this the same as HTML entities?
No. Percent escapes and ampersand references are different layers—see HTML Entity tool for the latter.
How this tool was checked
URL Encoder / Decoder was checked with spaces, plus signs, emoji, reserved characters, repeated encode and decode layers and invalid percent sequences. See the testing method or report a failure.