What is an IP Address?

An IP address is a network-layer address used to send data to and from a device, server, router, proxy, or other network interface. Your laptop may have a private address inside your home while websites see a public address supplied by your router, ISP, VPN, or proxy. That distinction is why one device can have several IP addresses—and many devices can appear online under the same one.

Quick Answer

An IP address, short for Internet Protocol address, is a numeric label assigned to a network interface so IP packets can be routed to and from it. IPv4 uses 32-bit addresses such as 192.0.2.34; IPv6 uses 128-bit addresses such as 2001:db8::34. An IP identifies a location in a network, not necessarily one physical device, person, or exact geographic location.

Key Takeaways

• IP addresses let networks route packets. Each IP packet contains source and destination addresses that help returning traffic reach the correct network path.
• IPv4 and IPv6 are the two versions in active use. IPv4 addresses are 32 bits; IPv6 addresses are 128 bits and provide a vastly larger address space.
• Public, private, static, and dynamic describe different properties. An address can be public and dynamic, private and static, or another valid combination.
• One IP does not always mean one device or person. NAT, CGNAT, proxies, VPNs, shared hosting, and changing assignments break that one-to-one assumption.
• IP geolocation is an estimate. An IP may suggest a network operator and broad location, but it does not contain a street address or a person's identity.
• A proxy or VPN changes the route and the address a destination sees. It does not erase cookies, accounts, browser signals, or the IP assigned to the underlying connection.

IP Address Types at a Glance

These labels describe separate dimensions and should not be treated as mutually exclusive categories:

For example, a home router can have a dynamic public IPv4 from its ISP, every laptop behind it can have a dynamic private IPv4, and all of them can share the router's public exit through NAT. A server, meanwhile, can have both a static IPv4 and one or more IPv6 addresses.

Methodology: We reviewed the existing Proxidize page and its linked content, current Google Search Console queries, IETF and IANA addressing standards, current Microsoft, Apple, and Android instructions, MaxMind's IP-geolocation guidance, and current Proxidize product pages on September 3, 2026. We did not test every operating-system release, router interface, carrier, or ISP assignment policy. Menu names and network behavior can vary by device and provider.

What Is an IP Address?

An Internet Protocol address is a logical address associated with an interface participating in an IP network. “Logical” matters: unlike a serial number stamped onto hardware, an IP address can be assigned, removed, leased, translated, shared, or replaced as a device moves between networks.

The Internet Protocol carries data in packets. An IPv4 or IPv6 packet includes a source address and a destination address. Routers inspect the destination and use routing tables to forward the packet toward the relevant network. A response addresses the observed source—or relies on intermediary translation state—to travel back; it does not have to cross the same routers in reverse.

The postal-address analogy is useful up to a point: the destination IP helps networks decide where to deliver a packet, and the source IP helps a response find its way back. But an IP address is not a deed to one device or home. Consider these common cases:

• a phone has different addresses on Wi-Fi and mobile data;
• a laptop can hold IPv4, IPv6, link-local, and VPN addresses at the same time;
• several devices can share one public IPv4 through a router;
• several subscribers can share one public IPv4 through carrier-grade NAT;
• a website can use several addresses or share one address with many other sites;
• a proxy can present its own exit address to the destination.

An address therefore needs context: which interface, address family, network scope, path, and moment? “What is my IP?” can have more than one correct answer.

How Do IP Addresses Work?

When you open a website, the browser normally starts with a domain name, not an IP address. The Domain Name System looks up one or more addresses associated with that name. The device can then send traffic toward an appropriate destination address.

Here is a simplified IPv4 path. Every public address in the example is reserved for documentation:

The flow has six main steps:

1. Name resolution finds a destination. DNS can return an IPv4 record, an IPv6 record, or several candidate addresses. Applications can also connect directly to an address.
2. The device chooses a source address and route. Its routing table determines whether the destination is local or should go through a router, VPN, proxy, or another gateway.
3. The application data is placed into packets. The IP header carries source and destination addresses. TCP, UDP, and QUIC add transport behavior above the IP layer.
4. Routers forward by network prefix. Internet routers do not need the page URL. They compare the destination with advertised prefixes and select the most specific matching route, a process described by CIDR routing rules.
5. An intermediary may change the visible source. A home NAT, carrier-grade NAT, forward proxy, or VPN endpoint can cause the destination to see a public egress address rather than the device's local address.
6. The response returns along a routable path. Stateful translators use their mapping tables to deliver it to the correct private device and connection.

The internet can choose different intermediate paths as routing changes. The address identifies the source and destination at the IP layer; it does not prescribe every router the packet must cross.

What Is an IP Address Used For?

What Does an IP Address Look Like?

The two IP versions in active use have different address lengths and text formats. Both are numbers underneath their human-readable notation.

IPv4 Address Format

IPv4 uses a 32-bit address, usually written as four decimal octets separated by periods:

Each octet represents eight bits and can range from 0 to 255. A 32-bit field has 2^32, or 4,294,967,296, possible bit patterns, although many blocks and individual values are reserved or unavailable for ordinary public-host assignment.

You may also see a prefix length:

The /24 means the first 24 bits form the network prefix. It does not mean the network portion is always the first three octets, nor that the last octet is always a device ID. Prefixes can end at other bit positions—for example, /20 or /27—so the CIDR prefix or subnet mask determines the boundary.

192.0.2.34 is part of a block reserved by RFC 5737 for documentation. It is safe for examples and is not a working public endpoint.

IPv6 Address Format

IPv6 uses 128-bit addresses. They are normally written as eight groups of hexadecimal digits separated by colons:

Leading zeros within a group can be omitted, and one consecutive run of all-zero groups can be compressed with :: :

The double colon can appear only once in an address because the omitted group count must remain unambiguous. The exact text rules are defined in the IPv6 Addressing Architecture. 2001:db8::/32 is reserved for documentation, so this example does not identify a live device.

A 128-bit field has 2^128, or approximately 3.4 × 10^38, possible values. Not all are globally assignable, but the space is dramatically larger than IPv4's—not merely “trillions” of addresses.

Is a Port Part of an IP Address?

No. A port belongs to a transport protocol such as TCP or UDP and helps select a service or connection on the addressed host. In this example, the address and port are separate values:

IPv6 literals are placed in brackets when a port follows them in a URL:

Port 443 commonly identifies HTTPS, but an address alone does not tell you which applications are listening or whether a firewall permits access.

IPv4 vs. IPv6

IPv4 and IPv6 are not simply two ways to write the same address, and they are not directly interchangeable. Modern networks often use dual stack, giving interfaces both versions, while translation and tunneling technologies connect environments that do not share a version.

IPv6 does not automatically make a connection private, anonymous, encrypted, or secure. Likewise, IPv4 is not automatically slower. Real performance depends on routing, peering, DNS, implementation, and the end-to-end path. For proxy-specific trade-offs, see the detailed IPv4 vs. IPv6 proxy comparison.

Public vs. Private IP Addresses

A public IP address comes from globally coordinated address space and can be routed across the public internet when its network advertises the relevant prefix. A private IPv4 address belongs to a reserved block intended for private networks and is not routed across the public internet.

RFC 1918 defines three private IPv4 blocks:

Only 172.16.0.0 through 172.31.255.255 is private—not every address beginning with 172. Separate networks can reuse the same private address because it has meaning only within its routing domain.

A typical home router gives devices private addresses, then uses Network Address Translation to share a public IPv4. That is why a laptop can show 192.168.1.24 in its settings while an IP-checking website reports a completely different address.

Public does not mean “open.” A globally routable address can still sit behind a stateful firewall that rejects unsolicited inbound traffic. Private does not mean “secure,” either: devices on a private network still need access controls, updates, segmentation, and appropriate firewall policy. The public vs. private IP guide explores the distinction in more depth.

Does IPv6 Have Private Addresses?

IPv6 does not use the RFC 1918 blocks. It has different scopes:

• Global unicast addresses most closely fill the public-address role.
• Unique Local Addresses use fc00::/7; locally assigned prefixes commonly begin with fd. They are intended for controlled internal routing, not the global internet.
• Link-local addresses use fe80::/10 and operate only on one link. Every IPv6 interface uses link-local addressing for important local functions.

An interface can have a link-local address, a global address, and temporary global addresses simultaneously. Seeing several IPv6 values on one device is therefore normal.

Private, Shared, and Other Special-Purpose Addresses

An IPv4 address outside RFC 1918 is not automatically a normal public address. The IANA special-purpose registry records several blocks with narrower meanings:

100.64.0.0/10 is particularly easy to mislabel. RFC 6598 calls it Shared Address Space, not a fourth RFC 1918 private range. If your router's WAN address falls inside it, your ISP may be using carrier-grade NAT. Follow the CGNAT checking guide to compare the router address with the public address seen externally.

Static vs. Dynamic IP Addresses

A static IP address is intended to remain assigned until an administrator or provider changes it. A dynamic IP address is assigned under a policy that allows it to be leased, reclaimed, and reassigned.

Static does not mean permanent forever, and dynamic does not mean constantly changing. A DHCP client can receive the same address repeatedly, while an administrator can renumber a static server. DHCP itself supports automatic, dynamic, and manual allocation modes; it is not synonymous with “changes every day.” IPv6 can also use stateless address autoconfiguration, DHCPv6, or manual configuration.

Public/private and static/dynamic answer separate questions. All four combinations are possible:

• a static private address for a printer;
• a dynamic private address for a phone on Wi-Fi;
• a static public address for an allowlisted server;
• a dynamic public address for a home internet connection.

Shared vs. Dedicated IP Addresses

A shared public IP can represent many private devices behind NAT, many subscribers behind CGNAT, many websites on one server, or many customer connections leaving through a service. Those are different kinds of sharing.

A dedicated address is reserved for one customer or workload under the relevant service agreement. It may be static, but the terms are not exact synonyms: dedicated describes who can use it, while static describes expected assignment stability. The dedicated vs. shared IP guide explains the operational trade-offs.

How to Find Your IP Address

First decide which address you need:

• Public IP: the egress address an external service sees for the current route.
• Private or local IP: the address assigned to an interface inside the current network.
• Router address: the default gateway your device uses locally.
• Router WAN address: the address assigned to the router on its provider-facing side.

These can all be different. You may also have both IPv4 and IPv6, plus addresses belonging to VPNs, containers, virtual machines, and inactive interfaces.

How to Find Your Public IP Address

Search for “what is my IP” or open a reputable IP-checking service. The result is the public source address used for that request. On a system with cURL installed, you can make a simple HTTPS request:

The result can differ by browser, app, or command if one uses a proxy, VPN, private-relay feature, or different IP version. An IPv4 result and an IPv6 result are both valid public addresses for their respective paths.

How to Find an IP Address on Windows

In Windows 10 or 11:

1. Open Start → Settings → Network & internet.
2. Select Wi-Fi and the connected network, or select Ethernet.
3. Look under Properties for the IPv4 and IPv6 addresses.

Microsoft documents the current path in its Windows network-settings guide. You can also open Command Prompt and run:

Use the active adapter's address. Entries for disconnected, virtual, VPN, or container adapters may not describe your current route.

How to Find an IP Address on macOS

Choose Apple menu → System Settings → Network, select the active connection, and open Details. For Wi-Fi, Apple lists the current IP and router addresses in the connection details. You can also Option-click the Wi-Fi icon in the menu bar to inspect connection data, as described in Apple's Wi-Fi guide.

That screen normally shows the local address. Use an external IP check when you need the public egress.

How to Find an IP Address on iPhone or iPad

Open Settings → Wi-Fi, then tap the information button beside the connected network. The IPv4 Address section shows the device's local address, subnet mask, and router. IPv6 addresses may also appear.

This is the Wi-Fi interface's address, not necessarily the public address used through your router. For cellular or public egress, use an external IP check from the relevant browser or app path.

How to Find an IP Address on Android

On stock Android, open Settings → Network & internet → Internet, then select the connected Wi-Fi network and view its details. Manufacturer labels vary; some devices place the information under Connections, Wi-Fi, About phone, or Status. Google's Android Wi-Fi instructions note that menus vary by Android version.

As on iOS, the Wi-Fi details show a local address. A public IP check shows the current Wi-Fi, carrier, VPN, or proxy egress used by that request.

How to Find an IP Address on Linux

Most current Linux systems provide the ip utility:

Look for inet for IPv4 and inet6 for IPv6 on the active interface. Loopback (lo), container bridges, virtual adapters, and temporary IPv6 addresses can produce several results. hostname -I offers a shorter list on many distributions but can still return multiple local addresses.

Why Do I See More Than One IP Address?

Multiple results are normal. A device can have:

• one address per active network interface;
• IPv4 and IPv6 at the same time;
• link-local and globally routed IPv6 addresses;
• temporary IPv6 addresses used for outbound privacy;
• addresses for VPNs, containers, or virtual machines;
• a private address locally and a different public address after NAT;
• a proxy egress visible only to traffic configured to use that proxy.

The useful question is not “which value is the real IP?” It is which address applies to this interface, scope, protocol version, and traffic path?

What Can an IP Address Reveal?

An IP address contains no built-in fields for a person's name, street address, GPS coordinates, browser history, or passwords. Those details are not encoded in the digits.

External databases and network records can nevertheless associate a public IP or prefix with useful context:

MaxMind, a major geolocation provider, states that IP geolocation is not precise enough to locate a specific household, person, or street address. Mobile networks, VPNs, proxies, corporate gateways, anycast services, and provider routing can make the apparent location especially broad or misleading.

An IP can become more identifying when combined with account records, cookies, device signals, precise timestamps, or an ISP's assignment logs. Under CGNAT, a public address alone may not even distinguish subscribers; the source port, protocol, and timestamp can be necessary to identify a mapping. Treat IP logs as potentially sensitive operational data, but do not present an IP alone as proof of a person.

Can Someone Trace Your IP Address?

A service you connect to can normally see and log the public source address that reaches it. A registry lookup can identify the network responsible for the address range, and geolocation can estimate an area. That is not the same as tracing the address directly to a person's home.

Mapping a dynamic or shared address to a subscriber usually depends on the provider's records and the relevant timestamp; access to those records depends on the provider and applicable legal process. The average website visitor cannot derive your exact identity or street address from the IP string alone.

Knowing a public IP also does not automatically give someone access to a device. Actual risk depends on whether services are exposed, how the router and firewall are configured, whether software is vulnerable, and what other information an attacker has. The guide to what someone can and cannot do with your IP separates common myths from realistic risks.

IP Address vs. MAC Address, Domain Name, and Port

What Is the Difference Between an IP Address and a MAC Address?

An IP address is logical and routable across IP networks. A MAC address is a link-layer identifier used on a local network segment. When a packet crosses a router, its IP destination usually remains the same until translation or another intermediary changes it, but the Ethernet or Wi-Fi frame is rebuilt for the next local link.

That is why an internet website normally sees a public source IP but not your laptop's Wi-Fi MAC address. Modern devices can also use a different private MAC address for each Wi-Fi network, further showing that a MAC is not always a permanent universal identity.

What Is the Difference Between an IP Address and a Domain Name?

A domain name is a human-readable name such as example.com. DNS can map it to one or more IPv4 and IPv6 addresses. The relationship is not one-to-one: one domain can use several addresses for resilience and geographic routing, and many domains can share one web-server or CDN address.

DNS answers can also change over time. The browser uses the address to establish the network connection, then sends hostname information at higher layers so shared infrastructure can select the right certificate and website.

What Is the Difference Between an IP Address and a Port?

The IP address gets traffic to the relevant host or interface; the TCP or UDP port identifies a service or connection there. A web server might listen on TCP port 443, while DNS commonly uses port 53. Ports do not increase the IP address space, and a closed or filtered port does not mean the address itself is invalid.

Can You Change Your IP Address?

Yes, but first distinguish changing the address assigned to an interface from changing the public egress a website sees.

Turning a router off and on does not guarantee a new public IP. The ISP can preserve the lease, use CGNAT, or assign an address under policies outside your control. Manually choosing a random public address in device settings will not give you ownership or a working route; use values assigned by the network administrator or provider.

Changing the public IP also does not reset everything a service knows. Cookies, logins, URL parameters, browser storage, TLS and browser signals, and account history can remain. IP changes should be part of ordinary privacy, testing, network administration, or authorized data-access workflows—not treated as permission to ignore service rules. For a practical comparison of intermediaries, read VPN vs. proxy and the guide to changing the address a destination sees.

An IP address is only one observable signal. The free Proxidize Browser Fingerprint tool shows the browser, display, graphics, hardware, language, time-zone, and other properties that can remain consistent after the network route changes.

How Proxies and VPNs Change the IP a Website Sees

A direct connection, a forward proxy, and a VPN use different paths:

After enabling a proxy or VPN, run the free Proxidize WebRTC Leak Test to compare the browser's normal exit address with any public addresses exposed through WebRTC. A different address is not automatically a leak—the tool also checks for dual-stack IPv4/IPv6 routes—but it can reveal when WebRTC and ordinary browser traffic leave through different networks.

A forward proxy accepts application traffic and makes or relays the outward connection. Only applications or protocols configured to use it necessarily follow that route. A VPN normally creates a tunnel and can route a broader set of device traffic, depending on split-tunnel and operating-system configuration.

Neither tool makes the underlying ISP address cease to exist. Instead, the destination normally sees the intermediary's public exit. The intermediary still receives a connection from the client side, so privacy depends partly on the provider, configuration, protocol, and logging policy.

Encryption is a separate question. HTTPS can keep web content encrypted end to end through an ordinary tunneling proxy, but HTTP and SOCKS proxy protocols do not inherently encrypt all client-to-proxy traffic. A VPN normally encrypts the tunnel between the device and VPN endpoint; traffic beyond that point still relies on application protocols such as HTTPS.

Rotating vs. Sticky Proxy IPs

A rotating proxy can select another eligible exit by request, connection, interval, session boundary, or provider-specific rule. A sticky proxy session attempts to keep the same exit through a related workflow. Rotation is useful between independent observations; stickiness is useful when several requests need consistent network context.

“Refresh the page” does not guarantee a different address. Browsers reuse connections, a page loads many resources, and provider semantics vary. The IP rotation guide explains the distinction between the event that triggers selection and the policy that chooses the next exit.

Common IP Address Misconceptions

Common IP Address Problems

Do not disable firewalls simply to make a test pass. Start by identifying the relevant address and scope, confirm the intended route, and change one variable at a time.

Is It Safe for Websites to See Your IP Address?

Websites need a return path to answer a normal direct request, so exposing a public source address to the service you contact is an ordinary part of internet communication. An IP is not a password. You generally do not need to panic because a site or server log contains it.

It is still useful operational data. Avoid publishing logs unnecessarily, keep routers and exposed services patched, use a firewall, remove unneeded port forwards, prefer HTTPS, and do not rely on an IP allowlist as the only authentication factor for sensitive systems. Organizations should limit access to IP logs and retain them only as long as their security, compliance, and operational needs justify.

If you want a destination to receive an intermediary's address instead of the direct ISP egress, use a reputable proxy or VPN that fits the scope. That shifts trust; it does not eliminate it.

Where Proxidize Fits

For legitimate web scraping, localized QA, price or SEO monitoring, AI data collection, and market research, a proxy supplies a controlled public egress between the application and destination:

The application keeps its own local and provider-assigned addresses. For traffic routed through the access point, the destination normally receives the selected proxy exit as the network-layer source.

Proxidize Residential Proxies provide residential exits across 195+ countries with country, city, and ISP targeting, rotating and sticky sessions, and HTTP, HTTPS, and SOCKS5 support. Proxidize Mobile Proxies provide mobile-carrier egress for workflows that specifically need mobile network identity and session control.

The proxy changes the network path; it does not grant permission to access a site, replace respectful rate limits, or erase browser and account state. Customers remain responsible for complying with applicable laws and third-party terms.

The Practical Meaning of an IP Address

An IP address is best understood as a routable, scope-dependent network label—not a permanent ID card for a device or person. IPv4 and IPv6 define its format; prefixes define network boundaries; assignment policy determines whether it stays stable; and NAT, CGNAT, VPNs, and proxies determine which source address another system actually sees.

When troubleshooting, always identify the layer first: public or local, IPv4 or IPv6, interface or gateway, direct or proxied, static or dynamic. That one habit resolves most of the apparent contradictions around “my IP address.”