Skip to content

Link Checker

See where a short or suspicious link goes, hop by hop, before you click it

How this check works

This page never opens the link. Our server asks each hop in the chain for its status and next address, once, without cookies, and the address is not stored or logged. It reads headers only, not the page, and it does not scan for malware.

Learn more

Guide and answers to common questions about Link Checker

What a short link hides

A short link such as bit.ly/3xYz or t.co/abc is only a pointer. The address you see says nothing about the website behind it, because the shortening service keeps the real destination in its own database and sends your browser there when you open it. You cannot tell a shop from a scam by looking at the text.

The same is true of many long, tidy-looking links in newsletters, ads and text messages: they often point to a tracking service first, which records the click and then forwards you to the page you expected.

How redirect chains are used for tracking and phishing

When a server answers a request with a redirect (status 301, 302, 303, 307 or 308), it tells the browser to ask somewhere else. Browsers follow these automatically and quickly, so a chain of several hops looks like a single click. Each hop is a different server that sees your request and can log it, along with your IP address, the time and whatever the link carries in its query string.

Marketing tools use this on purpose: a mail provider wraps every link in a tracking address, an ad network passes the click through its own server, and a shortener counts the visit before forwarding it. Attackers use the same mechanism for other reasons. A link that starts on a trusted domain and forwards to a fake login page passes a quick glance, and a chain of hops makes automated filters and human reviewers less likely to notice where it ends. A link can also start on https and be sent through a plain http hop on the way, which anyone on the network path can read or change.

Seeing the chain does not prove a link is good or bad. It shows where a click passes through, so you can judge whether the destination matches what you were told.

What UTM, fbclid and gclid do

Tracking parameters are extra pieces added to the end of an address after a question mark, for example ?utm_source=newsletter&fbclid=IwAR0.... They do not change the page you get. They tell the site, or the company that made the parameter, where the visitor came from.

  • utm_source, utm_medium, utm_campaign, utm_term and utm_content are campaign tags read by Google Analytics and many other analytics tools. They say which mail, post or ad you came from.
  • fbclid is added by Facebook and Instagram to links shared there. It is a click identifier that Meta can connect to the account that clicked.
  • gclid, dclid and msclkid are click identifiers from Google Ads, Google Campaign Manager 360 and Display & Video 360, and Microsoft Advertising. ttclid (TikTok), twclid (X) and yclid (Yandex) do the same for their platforms.
  • mc_eid and mc_cid (Mailchimp), _hsenc and _hsmi (HubSpot), vero_id, igshid, s_cid and similar names identify the recipient or campaign of an email or a shared post.
  • ref and referrer are often tracking and sometimes a real input, for example a branch name or a page to return to. This tool lists them as "possible" and does not remove them.

How this link checker works

You enter an address and our server requests it with a plain HEAD request, which asks for the status line and headers but not the page. If a server refuses HEAD, or answers a redirect without saying where, the server asks again with a GET that requests a single byte. When the answer is a redirect, it reads the Location header and repeats for the next address. It stops at a page that does not redirect, after 10 hops, or when a hop does not answer within 8 seconds.

The request identifies itself as ExposyLinkChecker/1.0, carries no cookies and no login, and never goes to anything but http or https. Before every hop the server resolves the host name and refuses to connect if any address it resolves to is a private or reserved one, such as your router's network, a loopback address or a cloud metadata address; it also refuses local names, unusual ports (only 80, 443, 8080 and 8443 are followed) and then connects to the addresses it checked. This is what keeps the checker from being used to reach into our own network.

The result is the chain, the final address, the tracking parameters found in any hop, a note when a hop is a well-known shortener, and warnings for patterns that deserve a second look: a hop that drops from https to http, more than 3 hops, a final site that differs from the one the link started on, a second shortener in the chain, an address with a user name in front of the @, a host name written with look-alike letters from other alphabets (shown decoded), and a very long query string.

How to read the chain

Each hop shows the host, whether it uses https or http, and the status the server gave. A 301 or 308 is a permanent move, a 302, 303 or 307 a temporary one; shorteners and trackers usually use these. A 200 is the end of the chain. A 404 or 410 at the end means the page is gone, and a 403 can mean the site turns away automated requests rather than that the link is broken.

Start with the last host. Does it match who sent you the link? Then look at what is in between. A tracker or shortener in the middle is normal for a newsletter and worth a second thought for a message from a bank. Any hop to plain http is a reason to be careful, and a link that ends on a different site from the one it started on, other than through a shortener, deserves a closer look at the final address.

The verdict at the top sums this up: good when the chain ends on https with no tracking parameters and no warnings; look at it first when there are tracking parameters or warnings; bad when a hop or the final page is plain http or the chain could not be followed to an end; and could not check when our server could not reach the host at all.

What this tool does not do

It does not scan the page for malware, does not look the address up in any reputation database or blocklist, and cannot tell you whether the site is a scam. A good verdict means the chain looked ordinary, nothing more: a freshly registered phishing site on https with no tracking parameters gets the same verdict as a bank.

It follows redirects that the server sends in its headers. A page that forwards you with JavaScript or an HTML meta refresh tag, or one that shows different things to different visitors, is not followed; the last hop can still lead somewhere else once the page loads. Some sites also answer our server differently from the way they answer your browser, for example by refusing requests that do not look like a browser. If a link only works for logged-in visitors, we see what an anonymous visitor sees.

What a clean link means

The clean link is the final address with the known tracking parameters taken out and everything else left exactly as it was. It points at the same page. It is useful when you share a link and do not want to pass on the campaign or click identifier attached to your own visit, or when you want to save an address without the noise. Parameters that might be a real part of the address, such as ref, are kept, because removing a real parameter can break the page. The clean link is not a safe link: if the destination is bad, the clean version of it is just as bad.

Privacy

The address you enter is sent to Exposy's server so that it can be followed, because a browser page cannot read the redirect chain of another site. The server uses it for this one check and does not store or log it, and the page does not open the link in your browser. The sites in the chain do see a request from our server, from our IP address and not yours.

Frequently asked questions

Is this link safe?

This tool cannot say. It shows where the link goes, whether the connection is https, and which tracking parameters and warning signs the chain has. It does not scan the page for malware and does not check any reputation database. If the destination is not what you expected, or the message asking you to click came out of the blue, do not open it.

Why is the destination different from the link I was given?

Because the link you were given is a pointer. Shorteners, email tools, ad networks and social sites send your click through their own server first, to count it and to forward you.

Does checking the link visit it?

Yes. Our server makes a request to each address in the chain, once, to read the status and where it points next. The request comes from our server, carries no cookies and is identified as ExposyLinkChecker. Nothing is sent until you tick the box above the button that says so. Your browser never opens the link. The check normally asks for headers only, so the page is not downloaded or run, but a link that acts when it is requested, such as an unsubscribe or confirmation link, can still act on it.

What is a tracking parameter?

It is an extra item in the query string of an address, such as utm_source, fbclid or gclid, that tells a site or an advertiser where a visitor came from or identifies the click. It does not change the page. Some of them, such as fbclid and gclid, are unique to the click, which lets the company connect your visit to an account or an ad.

What does "copy clean link" remove?

The known tracking parameters of the final address: any utm_ parameter, fbclid, gclid, dclid, msclkid, ttclid, twclid, mc_eid, mc_cid, igshid, yclid, _hsenc, _hsmi, vero_id, wickedid, oly_enc_id and s_cid. Everything else stays as it was, including ref and referrer, which are only marked as possible tracking. Other names that carry an identifier are not recognised.

Why can the tool not reach some links?

The host may be down or not exist, may take longer than 8 seconds to answer, may use a certificate our server does not accept, or may sit on a private address, which our server refuses to contact. In each case the result says which hop failed and why, and shows the chain up to that point. A failure to reach a host is shown as could not check, not as a bad link.

Why does the tool refuse some addresses?

Our server makes the requests, so it must never be tricked into reaching our own network. It refuses addresses that resolve to private or reserved ranges, local names such as localhost, .local and .internal, and ports other than 80, 443, 8080 and 8443. A public link has no reason to need any of these. If a redirect in the middle of a chain points to one, the chain stops there and the result says so.

Why is there a warning when the link changes to a different site?

When a link starts on one site and ends on another, you can no longer judge it by its first address. That is normal for shorteners, which is why no warning is shown when the first hop is a known shortener, but for a link that claims to come from your bank or a shop, a different final site is the thing to check.

What does a host name starting with xn-- mean?

It is how a name written with letters outside plain English is stored on the internet. Some letters from other alphabets look exactly like Latin ones, which attackers use to imitate well-known sites. The tool shows the decoded name so you can see what the browser would display, and warns about it.

Can I check a link from a text message or an email without opening it?

Yes, that is the use. Copy the link without tapping it (press and hold on a phone, or right-click on a computer), paste it here and read the chain. If the sender or the message is suspicious, go to the website yourself by typing its address.