About IPGet

What IPGet is

IPGet is a set of network diagnostic tools that answer questions about a connection: what address the internet sees, which network operates it, what a domain’s DNS actually returns, whether a certificate is valid, how fast the line is. Every tool is free, needs no account, and works in seven languages.

It exists because the obvious way to build these tools is to collect everything and worry about it later. IPGet is built the other way round: each feature starts from what it can avoid knowing, and the architecture makes that a structural property rather than a policy statement.

Who it is for

Mostly people who are already in the middle of a problem. Someone whose site stopped resolving after a nameserver change, an administrator working out why mail is being rejected, a developer checking whether a port is reachable from outside their own network, a person who wants to confirm their VPN is actually carrying traffic before they rely on it. The tools are built for that moment, which is why every result page explains how to read the result rather than only presenting it.

It is also used by people learning how the internet is put together. A diagnostic that shows real output from a real query is a better teacher than a diagram, provided somebody explains what the output means — so the explanations on each tool are written to be read by someone who does not already know the answer.

Why the site reports uncertainty

A diagnostic tool that always produces a confident answer is not more accurate than one that sometimes says "unknown" — it is just less honest about its inputs. The VPN detector returns yes, no or unknown, with a stated confidence, because network registration data genuinely does not always distinguish a consumer connection from a datacentre. A DNS lookup that fails is reported as a failure rather than as "no record", because "we could not check" and "there is nothing there" lead to opposite conclusions.

The same applies to what the tools decline to claim. IP geolocation is presented as an estimate with the country far more reliable than the city, because that is what the underlying data is. A port check reports what one server observed from one network path, not a universal fact. An SSL certificate can be fully readable and completely untrusted at the same time, and the result says both. Being specific about limits is more useful than implying there are none.

What IPGet does not do

It does not identify people. Every tool here describes a connection, a domain or a browser — never a person — and no result should be treated as evidence about an individual. It does not keep a history of what anyone looked up: queries are answered and discarded, so there is no log to hand over, sell or leak. It does not require an account, because a free diagnostic tool has no reason to know who you are.

It also does not sell recommendations. The site carries advertising, and that is disclosed, but no diagnostic result is influenced by it — the VPN detector in particular is deliberately kept away from any commercial relationship with a VPN vendor, because a verdict that could be influenced by who pays is worth less than no verdict at all.

How the tools are maintained

Every tool is covered by automated tests that run before any deployment, including tests that assert the things a reader cannot check for themselves: that private and reserved addresses are refused, that no query is written to storage, that the content security policy carries no wildcards, and that no page claims a rating, review or credential that does not exist. When a limitation is found, the usual fix is to document it on the page rather than to hide it.

The reference data behind the tools — the MAC address vendor list, the port and record-type tables — is derived from published registries rather than compiled by hand, so it can be refreshed rather than slowly drifting out of date. What each tool actually measures, and what it cannot, is written up in full on the how-it-works page.

How privacy works here

• No accounts. There is nothing to sign up for and no identity to attach a history to.
• No stored addresses. Lookups are not logged against a visitor, and the server-side caches are namespaced so a cached result cannot reveal who asked for it.
• No cookies of our own, and no tracking pixels.
• Local storage only where you ask for it. The two history features write to your own browser and nowhere else, and neither stores an IP address — one keeps a truncated one-way hash instead.
• Analytics are on, and they carry a tool name and a language — never an address, a location, or anything you typed into a tool.

Client-side and server-side, and why the split matters

Browser facts — screen size, timezone, language, platform, User-Agent — are read locally and rendered locally. No request is made for any of them, so they never leave your device.

Network checks cannot work that way. Finding out whether a port is open, what certificate a host serves, or what a resolver returns requires a connection from somewhere other than your browser, which is exactly what makes the answer meaningful. Those checks run on IPGet’s server, and the pages say so where it matters.

Security approach

Every check that accepts a destination validates it before connecting: the address is resolved, every returned address is checked against the private, reserved and cloud-metadata ranges, and the connection is made to the pinned address rather than re-resolving the name. That closes DNS rebinding as a class rather than as a case.

The site itself ships a strict Content-Security-Policy with no unsafe-inline, no unsafe-eval and no wildcard sources. Every script is bundled and hashed, which is why features that normally need an inline snippet are implemented as modules instead.

Technology

Open standards, not a black box

Everything here is built on published standards — RFCs for addressing, DNS, TLS and HTTP — and the tools report what those standards define rather than a proprietary score. Where a result is uncertain, the page says so: an unknown is reported as an unknown rather than rounded into a confident answer.