Normalise and compare URLs with every change listed and marked as meaning-preserving or not, so you can tell which ones are safe to apply and which are not.
Waiting for a URL
The safe options are on. The ones that can change meaning are off.
Waiting for a URL
Paste a URL to see what normalising it would change.
URL Normalizer runs entirely on your device — nothing is uploaded, and no URL is fetched.
You’re on the Free plan. Compare plans
Most URL normalisers apply a bundle of changes and hand back a string. The useful question is which of those changes you can actually afford, and that is what this reports: every transformation is listed, and each one is marked as meaning-preserving or not.
Lower-casing the host and resolving `.` and `..` are safe — the scheme and host are case-insensitive by specification, and a parser resolves dot segments anyway. Removing a trailing slash is not safe: `/dir` and `/dir/` are different resources, and many servers redirect between them rather than treating them as one.
Removing an empty parameter is not safe either, and the reason is easy to miss. `?debug=` is often a flag that is on, and `?q=` can mean an explicit empty search rather than no search at all. An absent parameter and a present empty one are different inputs.
Removing tracking parameters is reported as not meaning-preserving, which is stricter than most tools and is the honest answer. The URL still reaches the server: server-side conversion APIs consume `gclid` and `fbclid`, referral programmes credit a partner from `utm_campaign`, and onboarding flows branch on `utm_source`. It is usually harmless, which is not the same as safe.
The comparison view takes two URLs and shows what differs, parameter by parameter, which is the quickest way to find out why one of them works and the other does not.
No. `/dir` and `/dir/` are different URLs and may be different resources. Many servers redirect one to the other, which means removing the slash can turn one request into two.
Usually, and not always — which is why this reports it as not meaning-preserving. The parameter still reaches the server, and server-side conversion tracking, referral attribution and onboarding flows all read these values in real systems.
Because where a parameter repeats, a server that reads the first occurrence sees a different value after sorting. Sorting is what makes two URLs comparable as cache keys, so it is often worth doing deliberately.
That every change applied is guaranteed to preserve meaning, so the result can be substituted for the original. If any applied change is not, the flag is false — it is not an average or a score.
No. Nothing is fetched. A redirect chain can only be found by making requests, which this does not do.