URL Normalizer
Different-looking URLs often point to the same page. Paste a list and the normalizer rewrites each into one consistent form: the safe RFC 3986 normalisations are always applied, and optional rules let you force https, drop “www.”, remove index files, tracking parameters and fragments, and sort query parameters. Duplicates that appear after normalising are removed.
- Runs in your browser
- No sign-up
- Free to use
| Original | Normalised |
|---|
How to use URL Normalizer
- Paste URLs, one per line.
- Choose the optional normalisations that suit your site.
- Compare each original with its normalised form.
- Copy or download the clean, de-duplicated list.
URL Normalizer features
RFC 3986 normalisation
Case, default ports, dot segments and percent-encoding.
SEO rules
https, no www, no index files, optional trailing slash.
Tracking removal
utm_*, fbclid, gclid, msclkid and similar parameters.
Stable queries
Parameters sorted, empty ones removed.
De-duplication
One line per distinct normalised URL.
Before and after
A table shows every change.
When to use URL Normalizer
- De-duplicating URL lists from crawls and analytics.
- Preparing URLs for comparison or a redirect map.
- Cleaning tracking parameters from shared links.
- Choosing a consistent format for canonical URLs.
URL Normalizer FAQ
Which normalisations are always safe?
Those defined by RFC 3986: lower-casing the scheme and host, removing default ports, resolving /./ and /../, decoding unreserved characters such as %7E and upper-casing other escapes. They never change which resource is meant.
Why are the other rules optional?
Because they depend on the site. Most sites serve the same page with and without www or with and without index.html, but not all. Check before applying them to your own URLs.
Is sorting query parameters safe?
For almost every web application, yes. A few applications depend on parameter order; untick the option for those.
Which tracking parameters are removed?
utm_*, fbclid, gclid, gbraid, wbraid, dclid, msclkid, mc_cid, mc_eid, igshid, yclid and several others used only for analytics.
Does it check that the URLs work?
No. It works on the text only.
Is anything sent?
No.
One page, one URL
A single page can appear under dozens of addresses: with upper-case letters in the host, with an explicit :443 port, with index.html at the end, with tracking parameters from every campaign. To a computer comparing strings, each is different. Normalising rewrites them into one form so that equal pages compare equal.
RFC 3986, the URL standard, defines a set of normalisations that never change meaning. Scheme and host are case-insensitive, default ports are redundant, dot segments are shorthand, and percent-encoding of unreserved characters is optional. Applying these is always safe and is the first step of every crawler’s de-duplication.
Beyond that, sites follow conventions. Most choose either www or the bare domain and redirect the other, serve index pages at the directory URL and ignore analytics parameters. Applying the same choices to a list of URLs mirrors what the site does and collapses the variants.
Normalised lists are easier to compare, cheaper to crawl and better material for redirect maps and canonical tags. Keep the original list as well, because some variants may still need redirects to the normalised form.