Enter a base URL and as many key/value rows as you need, and get a correctly encoded URL as you type. Encoding is done by the standard implementation rather than a hand-written escape list, so the awkward characters — ampersands, equals signs, hashes, quotes, anything non-ASCII — are handled properly. One choice is worth understanding, because it causes a bug that is genuinely hard to spot: a space can be written as %20 or as a plus, and those are not equivalent. A plus means a space only under form-encoding rules. An API reading the query as a generic URI sees a literal plus, so your search for 'hello world' quietly becomes a search for 'hello+world' and returns nothing, while the URL looks perfectly fine in the address bar. The default here is %20, which means a space everywhere. Paste a URL that already has a query and its parameters become editable rows rather than being lost.
| Value | Encoded |
|---|---|
| hello world | hello%20world (or hello+world) |
| a+b | a%2B — a real plus, not a space |
| a&b | a%26b |
| a=b | a%3Db |
| a#b | a%23b |
| café | caf%C3%A9 |
| (empty) | key= — kept, not dropped |
No. The URL is assembled in your browser and never requested.
%20 unless you are posting to a form handler. A plus means a space only under application/x-www-form-urlencoded rules; anything treating the query as a generic URI reads it as a literal plus. That mismatch produces a search that silently returns nothing while the URL looks correct.
It is escaped to %2B before any space conversion happens, so it survives in either mode. That is checked by the tests in both settings, because getting it wrong would corrupt real data rather than just formatting it oddly.
It is kept, and its parameters are split into editable rows. Appending a second query string to a URL that already has one is a common way to produce a URL where the later parameters are ignored.
Yes — each row has a toggle. Useful for working out which parameter is causing a problem without retyping the others.
Yes. Repeating a key is legal and meaningful to some APIs, which read it as a list. They are emitted in the order you enter them.
Parameters are assembled with URLSearchParams, so every value is percent-encoded according to the standard rather than by a hand-written escape list that will eventually miss a character. The interesting detail is how a space is encoded, because a query string has two legal answers and they are not interchangeable. URLSearchParams always writes a plus, following the application/x-www-form-urlencoded rules that HTML forms use. But a plus only means a space under those rules: anything reading the query as a generic URI — a great many REST APIs, and anything treating the value as part of a path or a signature — sees a literal plus character instead. That produces a bug which is invisible in a browser address bar and obvious only once a search returns nothing. The default here is therefore %20, which means a space in every context, with the plus form available for genuine form handlers. The conversion is safe in both directions because a literal plus a user types is already escaped to %2B before the substitution happens, so swapping the remaining plus signs cannot corrupt real data — a point the tests check explicitly in both modes. A query already present on the pasted base URL is preserved and its parameters split into editable rows rather than being silently discarded, which is what happens if you naively append a new query string to a URL that already had one. Rows can be disabled without deleting them, so a parameter can be toggled while testing, and rows with no name are skipped rather than emitting a stray ampersand. Everything runs in your browser.