Skip to content
MyIP.dog

IP fundamentals

What is an IP address?

Understand public and private addresses, IPv4 and IPv6, and what an address can actually tell you.

Published:

How an IP address delivers network traffic

An IP address is a numerical identifier used when devices send traffic across networks. It helps network equipment decide where a packet should go and where a reply should be directed. The letters IP stand for Internet Protocol. An IP address is therefore part of communication, not a permanent personal identity, a device's serial number, or a password that automatically grants access to its owner.

When you open a website, the destination receives a connection with a source address. That source might belong to your home provider, mobile network, employer, VPN, or privacy relay. The homepage shows the public IP address observed for the browser's connection. What it sees can differ from the local address shown in your device's Wi-Fi settings, and that difference is often expected.

Public and private scopes

A public IP address belongs to globally managed address space and can participate in internet routing when appropriately assigned and routed. A private address is intended for use inside local networks. Many homes can use the same private numbers independently because their routers keep the local networks separate. Public visibility and actual reachability are still different: a firewall can block incoming connections to a publicly addressed device.

The familiar 192.168.1.20 might be a laptop's private IP address. Its router can translate outbound connections so internet services see another source shared by the household. This process is commonly called network address translation, or NAT. A remote website ordinarily cannot inspect every local assignment merely because one device behind the router opened a page. Local network settings answer a different question from the browser's public observation.

Some providers add another translation layer, called carrier-grade NAT, which can share a public IPv4 address across customers. A single IP address can therefore represent many devices and people. That matters when interpreting rate limits, abuse reports, and incoming connection problems. It also explains why identifying a registered network does not establish which person generated an individual request at a particular moment.

Read IPv4 and IPv6 notation

IPv4 uses 32-bit addresses commonly written as four decimal numbers separated by dots. Each number represents one eight-bit part. IPv6 uses 128-bit addresses written in hexadecimal groups separated by colons. The two families have different address spaces and operational characteristics. Your device may support both, but an individual connection still uses a particular path and address family.

An IPv6 IP address can look different in two interfaces while representing the same number. Leading zeroes within groups can be omitted, and a run of zero groups can be compressed with ::. The IP lookup tool normalizes supported input so comparisons are easier. It deliberately rejects ambiguous legacy IPv4 forms rather than relying on interpretations that can vary between software systems.

Seeing an IPv4 IP address on one page does not prove that IPv6 is unavailable. A dual-stack website may be reached over either family. Use the IPv6 test to request separate IPv4-only and IPv6-only endpoints. A successful request is evidence of that path; a failure still needs interpretation because filtering, DNS, or endpoint availability can also affect it.

Why one device can have several assignments

A device may have separate Wi-Fi, Ethernet, cellular, tunnel, and virtual interfaces. It can also hold addresses with different scopes and lifetimes. The IP address selected for one connection depends on routing and application behavior. A browser using a proxy may appear different from a command-line program that connects directly. No single website can inventory every interface and every application's policy simply by receiving one request.

IPv6 devices can use temporary source addresses, and private IPv4 assignments can change through DHCP. A changing IP address is not automatically a sign of compromise. Compare the network and connection conditions before drawing conclusions. If you need a stable destination for remote access, the correct solution depends on provider service, routing, firewall policy, and sometimes a naming service rather than a one-time observation.

Static and dynamic assignment

A static IP address is assigned with the intention that it remain stable under the relevant service or configuration. A dynamic assignment can change as leases, sessions, or networks change. One value remaining the same for weeks does not prove that the provider guarantees it as static. Read the service terms or ask the provider when stability is operationally important.

Restarting a router does not guarantee a new IP address. The provider can return the same lease or use assignment behavior that is not controlled by the customer. A VPN can change the exit for traffic it carries, while switching to another network can change the public source. Change one condition at a time when diagnosing behavior so you can connect the result with the action that caused it.

What the number can and cannot reveal

A registry lookup can describe the network range containing an IP address and identify available allocation information. A DNS PTR record can attach an operator-selected name. A geolocation database can estimate a country, region, or city. These are separate sources with different meanings. None should be silently substituted for another merely because a result page has an empty field to fill.

An IP address does not reliably reveal an exact home location or a person's name through an ordinary public lookup. At the same time, it is not meaningless for privacy: services can associate connection information with accounts, logs, and other data they already hold. Share it thoughtfully, particularly in a report that also contains timestamps, internal hostnames, credentials, or unrelated personal details.

An address alone is not proof of safety or malicious intent. A hosting network may contain ordinary services as well as abusive activity. A residential network can include a compromised device. Evaluate observed behavior and relevant evidence instead of treating an IP address category as a definitive reputation score. Investigations require more than a plausible organization name or map marker.

Ports and domain names

A domain name is a human-friendly label that can resolve through DNS to one or more addresses. A port identifies an endpoint used by transport protocols and applications on a host. An IP address, domain, URL path, and port are related concepts but not interchangeable inputs. If a tool asks for a domain, pasting a full URL can make the query fail or ask an unintended question.

For example, https://example.com/help contains a scheme, hostname, and path. DNS can help obtain destination addresses for the hostname, but it does not resolve the /help path. A subnet prefix such as /24 describes a network range rather than a website path. Use the subnet calculator when an IP address is accompanied by a CIDR prefix and your task concerns the range boundaries.

A practical way to use this information

When support asks for an IP address, confirm whether it needs your public source, a device's local address, the router's gateway, or a server destination. These values can all be correct while being different. Record the active network, address family, time, and VPN state. A precise description prevents a simple terminology mismatch from turning into an unnecessary configuration change.

If a website fails, first identify the layer involved. DNS might fail to return a destination, a connection might be blocked, encryption negotiation might fail, or the application itself might return an error. Knowing the IP address helps with some of those checks, but it does not solve every layer. Preserve successful observations and investigate the remaining uncertainty rather than replacing working settings without evidence.

References and further reading

The Internet Protocol model is described in RFC 791, with IPv6 addressing in RFC 4291. RFC 1918 defines private IPv4 space, and RFC 6598 defines shared address space. Continue with public vs private IP, IPv4 vs IPv6, or static vs dynamic IP according to the IP address question you need to answer.

What is an IP address? — common questions

Is a network address the same as a device serial number?

No. It is used for communication and can depend on the interface, network, assignment policy, or intermediary. Devices can hold several values, and several devices can share an externally observed source.

Why does a website see another value from the one in Wi-Fi settings?

A router, provider, VPN, or relay can change the source visible to the destination. Local settings and a remote observation therefore describe different parts of the connection and can both be correct.

What do the dots and colons tell me about the family?

Dotted decimal notation is used for IPv4, while IPv6 uses hexadecimal groups separated by colons. They are different address families, not short and long versions of an interchangeable personal identifier.

Does a public assignment guarantee that incoming connections work?

No. Appropriate routing, firewall policy, and an application listener are still required. Public scope describes the address context rather than a completed test of every service that might use it.

Can a private value be reused in somebody else's home?

Yes. Private ranges are intended for independent local use. The same local number on separate networks does not establish that the devices are identical or that a public registry can identify either household.

Why does a service need a port as well as a host value?

The transport and port help identify the intended communication endpoint at the host. A reachable network destination alone does not establish which application is listening or whether it will accept the request.

Is a domain name permanently tied to one numerical destination?

No. Naming records can change or return several addresses and families. Keep the resolver result and time when comparing a hostname with a network observation rather than assuming a permanent one-to-one identity.

How do I choose between local settings and the browser's result?

Start with the task. Local accessory configuration usually concerns the relevant interface and network; an internet service's source check concerns what that destination observes. Keep the method and family beside the copied value.