Separate the address from the sales pitch

Read a URL and Find the Real Domain

An invoice message can show a familiar company name while pointing somewhere else. Reading the whole address as one sentence makes that easy to miss. Instead, separate the hostname from the surrounding text, then work out the domain boundary. In this guide, real domain means the domain actually present in the supplied URL. It does not mean a verified company, a genuine invoice, or a destination that has passed a safety check.

Open the Link Inspector with one complete address you want to understand. The tool reads URL structure locally and presents plain-text results, never active destination links. Start with the hostname and registrable domain before revealing the rest. A sender can place reassuring words in many other parts of an address, so the order in which you read matters more than how familiar the first few words look.

Begin with one exact address, not the message around it

Suppose a supplier sends a button labelled View invoice. The button label is not the address. Use the message application to copy the link address without opening it, where that option is available. Paste that address into the inspector and choose Inspect link. Do not include the greeting, quotation marks, an HTML tag, or several links at once. You are trying to answer a narrow first question: what address did this particular button contain? The wording around the button is context, not part of the hostname. If the message has separate View invoice and Contact support buttons, inspect their addresses separately. Similar labels or matching artwork do not establish that the two buttons name the same host.

The input must be one absolute HTTP or HTTPS URL, including http:// or https://. A relative path such as /invoice/42 has no independent hostname and is rejected, as are custom schemes. Controls, bidirectional formatting characters, backslashes, and internal spaces are also rejected. Ordinary ASCII spaces outside the address are trimmed with a finding that tells you this happened. Other unexpected characters should prompt you to obtain a clean source address, not keep editing the text until the tool accepts something different.

There are three separate input limits: 8,192 Unicode code points, 32 KiB of UTF-8 data, and 200 query fields. A limit message describes this tool, not whether a website exists. Do not split a long signed address into pieces and treat the results as an inspection of the original. Load example offers a way to learn the controls without using a real invoice link. Use Clear before moving from that demonstration to your own address, then inspect only the single value you intended to examine.

Locate the hostname before trusting familiar words

Consider https://billing.example.com/invoices/42. The hostname is billing.example.com; the path begins at /invoices/42. Now compare https://example.net/billing.example.com/invoices/42. Here the hostname is example.net. The familiar-looking name has moved into the path and cannot change which host the address names. Read the inspector hostname field before scanning the full address. This simple habit is especially useful when the message preview cuts off the right side of a long URL or displays only a descriptive label.

An at sign introduces another common reading mistake. In https://accounts.example.com@helpdesk.example.net/invoice, the text before the at sign is user information, not the hostname. The host is helpdesk.example.net. User information can contain credentials, so do not reveal or share it just to illustrate the mistake. A long familiar prefix does not become the destination because it appears first. Likewise, a company name in a query value or fragment is data within the address, not an alternative host selected by the inspector.

Read scheme and port separately from the hostname. For HTTP, the effective default port is 80; for HTTPS, it is 443. An explicit nondefault port such as :8443 changes the port named by the URL without changing its registrable domain. Default ports may disappear from the normalized spelling. An HTTPS scheme describes the requested connection scheme; it is not a safety verdict. No connection is made here, so the inspector has not checked a certificate, a successful TLS connection, or the service listening on that port.

Use the suffix boundary instead of counting two words

A hostname is divided into labels by dots, but the last two labels are not always the registrable domain. In a teaching address such as portal.example.co.uk, the public suffix is co.uk and the registrable domain is example.co.uk. The label portal sits to its left. The inspector uses tldts with Public Suffix List data, including private suffix rules, to make this distinction. Read the reported boundary rather than inventing a universal rule based on the number of dots you can see.

Private suffix rules matter for hosted sites. For example, studio.github.io and gallery.github.io can be separated at the private suffix github.io, rather than both being grouped only as that platform name. This is a technical boundary in the supplied suffix data. It is not an ownership search, proof of who operates either page, or a business relationship between the two names. Private here describes a list section, not a promise that the website or its contents are private.

Keep the full hostname as well as the registrable domain in view. A subdomain can distinguish an account area from a support area, and identical domain boundaries do not make every page equally appropriate for your task. Conversely, an IP address is not a registrable domain, and some inputs have no useful suffix-derived domain result. A missing result does not establish maliciousness or nonexistence. The domain field is an interpretation of the supplied hostname using bundled rules, not a live registry lookup or a record of the current operator. When comparing two supplied links, check whether the full host changed, the boundary changed, or only the path changed. Those are different observations and can lead to different questions for the supplier.

Start with the host

Separate the address into readable parts

Inspect one complete HTTP or HTTPS URL and read the hostname before opening any masked details.

Reveal only the details needed for your question

Details are masked by default because the address may contain more than public routing information. Choose Reveal details only when you need to examine the normalized URL, path, ordered query pairs, fragment, or user information. A path can include a private document name, even when it contains no field called token. A fragment can also hold sensitive data. Hide details restores the masked presentation; it does not remove the information from the working session. Avoid revealing the address while screen sharing or capturing a screenshot for a general discussion.

For a harmless teaching example, examine https://example.com/invoice?item=red&item=blue&view=compact#totals. The path is /invoice, the query contains three pairs, and the fragment is totals. The two item pairs remain in their original order. Repeated names are not automatically a mistake, and the inspector does not decide whether an application uses the first value, the last value, or all values. Raw and decoded views help explain spelling; neither tells you how an unknown destination implements its own application logic. Keep a field name with its value when discussing a query. Saying that blue appears somewhere in the address loses the useful fact that it is the second value of a repeated item field.

A nested URL is still just a value. In https://example.com/continue?next=https%3A%2F%2Fexample.net%2Faccount, the outer hostname is example.com. Decoding the next value can make the embedded address readable, but it does not prove that the outer site redirects there. The value could be ignored, rejected, displayed, or used differently. Do not describe it as an observed second hop. If your task later becomes removing unnecessary attribution fields from a known publishing link, use the separate URL Cleaner and review its output as a new address.

Explain changed spelling without claiming a visit

The inspector uses the native URL parser to produce a normalized representation. This can lowercase a hostname, make an implicit path explicit, remove an explicit default port, or resolve dot segments in a path. For instance, https://EXAMPLE.com:443/a/../invoice can normalize to https://example.com/invoice. That is a parsing result, not evidence that a server moved the invoice. Keep a distinction between text normalization and a redirect. Only the former is available here; the normalized result is not a record of a browser journey. In a handoff, call it the normalized address rather than the address the site redirected to. That small wording choice prevents a colleague from assuming that a connection was tested.

Internationalized hostnames have ASCII and Unicode representations. An ASCII label beginning xn-- can represent a name containing non-ASCII characters, and the Unicode view can help you read it. The tr46 conversion used for that view is not a homograph detector. Do not conclude that a name is genuine because it displays neatly, or harmful because its ASCII version looks unfamiliar. When exact spelling matters, compare the ASCII hostname with an address obtained through an independent, trusted route, rather than relying on a logo or on visual similarity alone.

Findings call attention to address features; they are not a risk rating. An IP literal, credentials, an unusual port, or a normalization change may deserve an explanation in your context without proving abuse. Equally, an address with no surprising findings has not passed a reputation check. There is no fetch, API request, DNS lookup, status check, TLS test, or short-link expansion. If a short address hides a longer destination, this tool can explain the short address itself but cannot reveal the actual redirect chain or the page eventually shown.

Finish with a precise conclusion and a small handoff

Write the conclusion at the level the inspection supports. For the invoice example, a useful sentence is: the supplied address names helpdesk.example.net, while the company name appears in user information. That tells a colleague what you observed without claiming that you identified an attacker or established a legitimate supplier. If the question concerns an unexpected payment request, contact the supplier through an established channel. You can choose not to follow a message link without first producing a technical verdict about every part of it. Ask for the expected billing hostname or a replacement invoice through that existing relationship. Do not use contact details supplied only by the questioned message as independent confirmation of its own claim.

Choose Copy hostname when another person needs only the host. Choose Copy inspected URL only when the full normalized address is needed and its private contents are appropriate for that handoff. Copying is explicit; results do not become clickable links. Copy before changing tabs, because all working input and derived results are cleared when the page becomes hidden and on pagehide, pageshow, or freeze lifecycle events. Use Clear when finished as well. There is no saved history, project, report storage, PDF, or download from which to retrieve the inspection later.

The input runs in an isolated opaque sandbox, not in an ordinary WordPress input. Its no-store response and connect-none restriction limit retention and network access within that boundary. They do not erase clipboard contents or clipboard history, defeat browser extensions, or protect against a compromised parent that replaces the interface. Treat clipboard copies as separate copies under your control. A good stopping point is an understood hostname, an explained boundary, and a decision about the original task, without turning a private invoice address into a new shared document.

Questions about Find the Real Domain

Does the registrable domain identify the company behind a link?

No. It identifies a domain boundary derived from the hostname and Public Suffix List rules, including private rules. It does not identify a business, establish ownership, or confirm that the message came from that organization. Compare the address with information obtained independently when that relationship matters.

Why is the inspected URL different from the pasted text?

The native URL parser can normalize spelling, including hostname case, default ports, and path segments. Outer ASCII spaces are trimmed with a finding. These are local text-processing changes, not a destination visit or evidence of a redirect. Review the relevant finding before deciding whether the difference matters.

Can I return after a tab switch and copy the previous result?

No. Hiding the page clears its working state, and leaving, restoring, or freezing it also triggers clearing. Use the appropriate explicit copy control before switching tabs. The inspector has no recovery history, and clearing its state does not erase any copy already placed on the clipboard.

Continue with another link-inspection guide

Inspect a supplied address in the Link Inspector or choose a related task below. Explore more practical utilities in all free tools.

A clearer address, a narrower claim

Read the next link without opening it

Return to the free Link Inspector for hostname, domain-boundary, and URL-structure results in plain text.