What Is My IP

What is my IP — the address websites actually see, plus what your browser gives away

The large number at the top of this page is your public IP address: the address that every website, API and mail server sees when your connection reaches them. It is detected live from the request your browser just made, and it is not stored or logged — the lookup endpoint reflects the metadata that already rides along with your request straight back to the browser that made it, and returns nothing to anyone else.

Alongside the address, the page shows the approximate location and network owner derived from it, and a set of details read entirely inside your browser: operating system, browser and engine, device type, screen and viewport size, languages, time zone, and the raw user-agent string. The two halves come from different places, and knowing which is which is most of what people actually want to understand here.

What comes from a server and what stays local

The distinction is a hard technical one, not a design choice: JavaScript running in a page has no way to ask “what is my public IP?”. That answer only exists at the other end of the connection, so a server has to tell you. Everything else on this page is read locally and never sent anywhere. Refresh re-runs the lookup; Copy all puts a plain-text summary of everything shown on your clipboard, which is the format most useful to paste into a support ticket.

Public IP versus the 192.168 address on your own network

If you have ever looked in your operating system’s network settings and seen something like 192.168.1.42, and then found a completely different number here, nothing is wrong. They are two different addresses doing two different jobs.

Your router hands out private addresses to the devices in your home or office from ranges reserved for exactly this purpose in RFC 1918: 10.0.0.0/8, 172.16.0.0/12 and 192.168.0.0/16. Those addresses are not routable on the public internet; millions of networks use the same ones simultaneously without conflict. When your laptop talks to a website, the router rewrites the source address to the single public address your ISP gave it, remembers which internal device the conversation belongs to, and translates the replies back. That is NAT, and it is why the whole household appears to the outside world as one address.

Which address you need depends on the task:

• Allow-listing your office in a firewall, a VPN or an admin panel — the public IP shown above.
• Port forwarding a game server or NAS — both: the public address is what people connect to, the private address is what you forward to internally.
• Reaching a printer, a NAS or another machine on the same network — the private address, which this page cannot show you and which you find in your device’s own network settings or your router’s client list.
• Reporting an abuse complaint or a login alert — the public IP, since that is what the remote service logged.
• Confirming a VPN is actually carrying your traffic — the public IP, checked before and after connecting.

There is a third case that catches people out: on many networks the address on your device is a 169.254.x.x link-local address, which means it failed to get one from DHCP. That is a local fault, and this page will still show your public address only if some other path to the internet exists.

IPv4, IPv6, and why you might see both

An IPv4 address is four numbers separated by dots, such as 203.0.113.7. There are only about 4.3 billion of them and they ran out years ago. IPv6 addresses are eight groups of hexadecimal separated by colons, such as 2001:db8:85a3::8a2e:370:7334, and the space is large enough that scarcity is no longer the design constraint.

This page labels the version of whichever address your connection actually used to reach it. If you are on a dual-stack network — most modern mobile and fibre connections are — your device holds both an IPv4 and an IPv6 address, and which one gets used is decided per connection by the browser and the operating system. That is why the same device can show IPv6 here and IPv4 on a different lookup site, or flip between them across refreshes. Neither answer is wrong; they are both yours.

When the connection arrives over IPv6, a pseudo IPv4 row may also appear. This is a synthesised IPv4-shaped value provided so that legacy systems which can only store or match a four-octet address have something to work with. It is not an address you own, nothing routes to it, and it should not be used for a firewall rule that needs to be precise. If a service asks you to allow-list an address and you are IPv6-only, give it the IPv6 address, or ask whether it supports IPv6 at all.

One practical consequence of the two families: an allow-list built around your IPv4 address will silently fail the day your traffic goes out over IPv6, and vice versa. If a service supports both, register both.

What an IP does and does not reveal

An IP address is a routing label, not an identity document. What can be derived from it publicly is: the organisation that holds the address block, its autonomous system number, and a location estimate based on where that block is registered and observed to be used. That is genuinely useful for choosing a language, a currency or a nearby server, and it is roughly what advertising and fraud systems use it for.

What it does not reveal to a website is your name, your street address, your phone number or who is sitting at the keyboard. Your ISP knows which subscriber account held a given address at a given time, and that mapping is disclosed to third parties only through legal process. The gap between “a website can see your city” and “a website knows who you are” is wide, and the scare-tactic marketing around IP lookups routinely elides it.

Why the location is often wrong by miles

IP geolocation is inference, not measurement. There is no coordinate embedded in an address. Databases are built from registration records, routing announcements, latency measurements and ISP-supplied data, and the accuracy varies enormously by carrier and region. Common reasons the city shown above is not your city:

• The address maps to your ISP’s regional hub, not your street — often the nearest large city, sometimes a long way off.
• Mobile data routes through gateways that may sit in an entirely different region from the tower you are attached to.
• Business and satellite connections are frequently registered to a head office address rather than the site you are at.
• VPNs and proxies deliberately show the exit server’s location.
• Corporate networks may backhaul all traffic to a single internet breakout in another country.
• Reassigned blocks can sit in databases with stale registration data for a long time.

The coordinates shown here are labelled approximate for that reason: they typically indicate a city-level centroid, not a building. If you need a precise location, that comes from the browser’s Geolocation API with your explicit permission — a separate mechanism this page does not use and never prompts for. If a site is showing the wrong city or the wrong language, the fix is on their side or in your account settings, not something you can correct by changing your IP.

What changes your IP address

A static address is one your ISP assigns permanently, usually on a business plan and often for an extra fee. It is what you want if you are hosting anything, running a mail server, or if remote systems allow-list you. A dynamic address is assigned from a pool and can change — on most residential connections it is stable for long stretches and then changes without warning, typically after a router reboot, a line drop, a lease expiry or ISP maintenance. If the answer here differs from what you saw last month on a home connection, dynamic assignment is almost always the reason.

Things that change the address you see above:

• Connecting or disconnecting a VPN — the public address becomes the VPN exit node’s, and the network owner row will show the VPN provider rather than your ISP. This is the quickest way to confirm a VPN is genuinely routing your traffic.
• Switching from Wi-Fi to mobile data — a different carrier network entirely, often a different IP version too.
• Moving between networks — office, home, a café, a hotel.
• Rebooting the router or the modem, on a dynamic plan.
• ISP reassignment during maintenance or when a DHCP lease is not renewed.
• Browser privacy features that route traffic through a relay, which substitute a provider-operated address.

CGNAT: when you share a public address with strangers

Because IPv4 ran out, many ISPs — mobile carriers especially — place customers behind carrier-grade NAT, a second layer of translation above your own router. Hundreds or thousands of subscribers share one public IPv4 address. The symptoms are recognisable: port forwarding does not work no matter how correctly you configure the router, inbound connections fail, some services rate-limit or CAPTCHA you because of what a neighbour did, and a self-hosted service is unreachable from outside. Addresses in the 100.64.0.0/10 range are reserved for exactly this shared-transition purpose, so seeing one on your router’s WAN interface is a strong signal you are behind CGNAT. The usual remedies are to ask the ISP for a real public IPv4 address, use IPv6 if both ends support it, or put the service behind a tunnel that establishes the connection outbound.

Reading the rest of the page

The browser-side cards are worth a look because they show, concretely, what any website can read about your setup without asking permission — the user-agent string, the screen and window dimensions, the fonts of information that feed browser fingerprinting. Some of it is deliberately fuzzy: device memory is reported in coarse steps, and several browsers freeze or reduce the user-agent string precisely to make fingerprinting harder, which is why the OS version may read as a range rather than an exact release. The connection card will say the data is not exposed on Safari and Firefox, which do not implement the Network Information API; that is expected, not an error.

If the IP panel says the lookup is unavailable, the request to the endpoint failed — a blocked request, a captive portal or a dropped connection are the usual causes. The browser-side details will still be shown, because they never needed the network in the first place.