What Is My IP

The first time people check their IP they get confused: the router admin page shows 192.168.1.x, ipconfig on the command line shows the same thing, but a website reports a completely different address. The reason is that your device really has two layers of addressing. The ranges 192.168.x.x, 10.x.x.x and 172.16–31.x.x are RFC1918 private addresses (your internal IP) — valid only inside your home or office LAN and not routable on the public internet. The address that actually represents you online is your public IP, assigned by your ISP and globally unique.

The translation between the two is handled by NAT (Network Address Translation). Your home router keeps a mapping table: when an internal device reaches out, NAT swaps the source to the public IP and records the port, then restores it when the response comes back. This means a whole building, an apartment complex, or even an entire ISP metro network can share one public exit IP — that is carrier-grade NAT (CGNAT). It also explains why you and your neighbor can sometimes look up the exact same public IP.

The checker at the top of this page reads that public exit IP — the one any website, game server, or anti-fraud system actually sees as you. To confirm whether you have a dedicated public IP (which affects port forwarding, remote access, and self-hosting), compare the address on your router's WAN/dial-up interface with the one this page returns: if they match, it's usually a real public IP; if they differ, you are most likely behind CGNAT.

A complete IP picture is more than one string of numbers. IPv4 is the classic 32-bit address (like 203.0.113.7); its roughly 4.3 billion addresses ran out long ago, which is exactly why NAT sharing is so common. IPv6 is a 128-bit address (like 2408:8210:...) with a near-infinite supply, and most home broadband and mobile networks now run dual-stack by default. You may have an IPv4 exit and an IPv6 exit at the same time, and their ownership, risk score, and even whether they can reach a given site can all differ. Checking only IPv4 while ignoring IPv6 is a common blind spot that leaks your real location or trips risk controls.

Geolocation is not GPS — it's a database lookup. Providers map IP ranges to the locations ISPs register, then correct with network probing and user feedback. Be clear-eyed about precision: country-level is usually accurate, but published studies put city-level hit rates roughly between 20% and 75%, and accuracy within 10 km drops further. A city that's off by tens of kilometers, or one that resolves to the ISP's headquarters, is completely normal — never treat it as a street address.

ASN (Autonomous System Number) reveals whose network the IP belongs to — whether it sits in a telecom backbone, a cloud provider's data center, or a proxy vendor's range. Combined with the IP block's registration type, the ASN is the key signal for telling a residential network apart from a data-center / hosting IP — which is precisely what cross-border sellers and signup risk engines care about most. The checker on this page lists IPv4/IPv6, geolocation, ASN, and a residential-vs-hosting verdict together.

A common misconception is that once a VPN or proxy is on, the site can only see the proxy IP. That isn't true. The classic flaw is the WebRTC leak. To enable peer-to-peer audio/video, browsers use the STUN protocol to ask a server for their own public address — and that request travels over a separate UDP channel at the application layer. Many VPNs only capture network-layer traffic and can't intercept it. The result: a web page can run a snippet of JavaScript and grab the real IP your proxy was supposed to hide, with no prompt and no warning to you.

By default this affects Chrome, Firefox, and Edge; Safari and Brave ship leak protection in their default configuration. Beyond WebRTC, DNS leaks are another exposure path: if your domain-resolution requests bypass the proxy and go straight to your ISP's DNS resolver, an observer can infer your real ISP and rough region from where the lookups originate — even when your HTTP traffic genuinely runs through the proxy.

For cross-border operations, multi-account work, and privacy-sensitive users, seeing that the exit IP changed is nowhere near enough — you have to confirm there's no side-channel leak. The checker on this page runs both WebRTC and DNS leak detection and flags any real address that surfaces outside the proxy IP in red, so you catch the gap before signing up or logging in, instead of after an account ban.

Many people assume an IP just needs to "work," but risk engines never look at the address itself — they look at the reputation behind it. Within the same range, residential broadband and a data-center IP get completely different treatment: data-center IPs naturally score higher because they're so often used for scraping and bulk signups. If the IP has also appeared as a Tor exit, on a spam list like Spamhaus, or in an abuse database like AbuseIPDB, then captchas, step-up verification, or outright rejection are waiting for you.

A single data source is easy to get wrong — one database calling an IP clean doesn't make it clean. The more robust approach is to aggregate multiple sources for cross-validation and set a floor on "hard signals": if an IP matches Tor, a known proxy range, or an authoritative blocklist, the verdict can't score high no matter how good the other metrics look. There's also an overlooked dimension — the /24 neighbors (C-block profile): your own IP may be clean, but if the same /24 is packed with flagged abusive addresses, those bad neighbors drag down your real-world pass rate.

IPOK aggregates multiple risk sources — ip-api, ipapi.is, proxycheck, AbuseIPDB, Scamalytics, StopForumSpam, IPQS — and layers its own IPOK-DB (with Tor / X4BNet / Spamhaus data) as a hard-signal floor, plus the /24 neighbor profile and a residential-vs-hosting verdict. The IP you check is not logged. So this page doesn't just tell you "what my IP is" — it answers whether that IP looks clean to others and whether it's worth swapping.