URL Parser

The URL Parser takes any web address and splits it into the parts a browser sees: the scheme, the credentials, the host and port, the path, the query string, and the fragment. Instead of squinting at a long link full of percent signs and ampersands, you get a labelled breakdown and a decoded table of every query parameter, in the order they appear.

A URL is more structured than it looks. It follows the WHATWG URL standard - the same rules browsers use - with a defined grammar for each component. Reading it by eye is error-prone: a stray question mark, an encoded slash, or a repeated parameter is easy to miss, and those are exactly the details that break routing, tracking, or an API call. This tool uses the browser's own URL engine, so what you see is what the browser would parse.

Everything happens in your browser. Paste a link with tracking parameters, an API endpoint with a dozen query values, or a bare domain, and the parser resolves it locally. Nothing is uploaded, so it is safe to inspect URLs that contain tokens, session IDs, or internal hostnames. It works offline once the page has loaded.

Parsing follows the WHATWG URL standard, which is what browsers implement, and it is more forgiving than people assume: it will normalise a missing scheme, collapse dot segments in a path, and lowercase the host, because hosts are case-insensitive while paths are not. Two areas repay attention. Internationalised domains are stored as punycode, so a host in Cyrillic or Chinese displays in its own script and travels as an ASCII string beginning xn--, and those two forms are the same host. And percent-encoding is only meaningful in context: a slash encoded as %2F inside a path segment is data rather than a separator, which is exactly why decoding a whole URL in one pass and then re-parsing it gives the wrong answer.

A quick check of what this online URL parser returns: paste example.com/pricing?plan=pro&plan=team&q=trail+runner. It notes that https:// was assumed, lists the hostname as example.com with TLD .com, and shows port 443 marked as the scheme default. The Query parameters table has three rows, plan = pro, plan = team and q = trail runner, because a plus sign in a query means a space. The JSON copy button groups the repeated key as {"plan": ["pro", "team"], "q": "trail runner"}, ready to paste into a test.