What Is an IP Address? A Visual Beginner’s Guide to How Devices Are Grouped on the Internet

Have you ever seen numbers like these on your smartphone’s Wi-Fi settings or your home router’s admin page?

• 192.168.1.10
• 203.0.113.5
• 10.0.0.1

These are all IP addresses — the “addresses” assigned to every device connected to the internet (PCs, smartphones, servers, printers, smart home devices, and so on). The reason this page appears the moment you type “toolcluster.app” is that, behind the scenes, the string toolcluster.app is translated into an IP address, and that IP address is what tells data how to reach your device.

Now, 192.168.1.10 and 192.168.1.11 look similar as numbers — but do they belong to the same network? What about 192.168.2.10? Or 203.0.113.5? Once you can intuitively answer “how many leading octets need to match for two IPs to be in the same group”, you’ll be able to glance at any IP and instantly tell “that’s a home LAN address” or “that’s an external server.”

In this article, we’ll walk through, step by step:

• What an IP address is (the “address” role on the internet)
• Why the format is “four numbers separated by dots”
• How many leading octets must match to be in the “same network” — visualized in two views: a hierarchy tree and a color-coded table
• The difference between private IPs (like 192.168.x.x) and global IPs (your outward-facing IP)
• How to look up your own IP address

No advanced math, no specialist knowledge required. By the end, you’ll understand IP addresses from their building blocks up.

1. What Is an IP Address — The “Postal Address” of the Internet

1-1. The role of an “address” on a network

When you mail a letter, you need an address. Without something like 123 Main Street, Springfield, IL 62701, USA, the postal carrier can’t find your house.

The internet works in exactly the same way. Whether you’re streaming a video, sending an email, or opening a webpage, your data is split into small chunks of information called packets that travel between computers all over the world. Every packet carries both a destination address and a source address — and those addresses are IP addresses.

So an IP address plays two roles:

• An identifier assigned to every device connected to the internet
• An address that tells each packet “where it came from and where it’s going”

Just as no mail can be delivered without an address, no data can travel across the internet without an IP address.

1-2. Why four numbers — the 32-bit structure of IPv4

The most familiar IP address format looks like 192.168.1.10. This “four numbers separated by dots” format is the address layout of IPv4, the most widely used standard today.

Why four pieces? The answer is straightforward: IPv4 specifies that every address is 32 bits long (32 zeros and ones). Those 32 bits are split into four 8-bit groups, which is exactly what the four numbers in 192.168.1.10 represent.

In binary:    11000000 . 10101000 . 00000001 . 00001010
In decimal:        192 .      168 .        1 .       10
                  └1st─┘    └2nd─┘    └3rd─┘    └4th─┘
                   octet     octet     octet     octet

Each 8-bit chunk is called an octet. The word octet comes from Latin and means “a group of eight” — in computer networking, “8 bits = 1 octet” is the fixed definition (essentially the same as a byte).

1-3. Why each octet ranges from 0 to 255

Each octet of an IP address only takes a value between 0 and 255. There’s no such thing as 192.168.1.300. This is because of the range you can express in 8 bits.

With 8 bits you can represent 2^8 = 256 distinct patterns. Counting from 0, that’s the values 0 through 255 — exactly 256 numbers. That’s why every octet in an IP address fits in 0–255.

Putting that together, the total number of possible IPv4 addresses is:

256 × 256 × 256 × 256 ≈ about 4.3 billion

That sounds like a lot at first glance, but as we’ll see in §7, even 4.3 billion turned out to be insufficient for the modern internet — which is why the IPv6 standard was eventually born.

2. Octets — The Four Building Blocks of an IP Address

In §1 we mentioned that an IPv4 address is “32 bits split into four 8-bit groups.” In this section, we’ll look more closely at how each of those groups is named and arranged. Once this is clear, the main diagram in the next section becomes much easier to read.

2-1. Naming the 1st through 4th octets

Take 192.168.1.10 as an example. From left to right, the four numbers are called the 1st octet, 2nd octet, 3rd octet, and 4th octet.

You’ll also see them written as “first byte,” “first segment,” and similar terms in technical documentation — they all refer to the same thing.

2-2. Bigger groups on the left, individual hosts on the right — same as a postal address

The order of the octets carries an important meaning: moving left points to bigger groups, and moving right points to individual devices.

It helps to think of it as a postal address.

Address:    USA       California   San Francisco   123 Market St
            └Country┘ └──State──┘  └────City────┘  └──Street───┘
             Largest                                Individual

IP:         192       168          1               10
            └─────── Network part ───────────┘     └─Host part─┘

If two people share “USA” they’re at least in the same country; if they share “California” they’re closer; if they share “San Francisco” they’re neighbors. IP addresses behave the same way: the more leading octets that match (counting from the left), the “closer” two addresses are on the network.

Strictly speaking, this “address analogy” isn’t perfectly accurate. In real IP addresses, where the network part ends and the host part begins is configurable per address (CIDR notation, covered in §4). At this stage, just remember the directional intuition — bigger groups on the left, individual hosts on the right.

3. [Main Diagram] How Many Leading Octets Make Two IPs the “Same Network”?

This is the heart of the article. IP addresses that share their leading octets, like 192.168.1.10 and 192.168.1.11, belong to the “same group” on the network. By contrast, addresses that differ in the leading octets — like 192.168.1.10 versus 203.0.113.5 — sit on completely different networks.

We’ll explore this group structure through two views: a hierarchy tree and a color-coded table. Switch to whichever feels easier. The tree view also has a “Add colors / Remove colors” toggle, so whether you prefer plain monospace alignment or color hints, you can read it comfortably either way.

3-1. Visualizing “depth of grouping” with a tree and a colored table

The diagram below lines up three devices on the same home LAN, a PC on a different floor, and a PC on a completely different ISP. The tree view makes the “networks within networks” nesting obvious, while the color table makes “how many leading octets match” jump out at a glance.

192.0.0.0/8                          ← /8 network (broadest grouping)
└── 192.168.0.0/16                   ← /16 network
    ├── 192.168.1.0/24               ← /24 network (most common for home LANs)
    │   ├── 192.168.1.10             PC-A (home)
    │   ├── 192.168.1.11             PC-B (home)
    │   └── 192.168.1.50             Printer (home)
    └── 192.168.2.0/24               ← Different /24 (same /16 but different group)
        └── 192.168.2.10             PC on another floor

203.0.0.0/8                          ← Completely separate tree (different ISP)
└── 203.0.113.0/24
    └── 203.0.113.5                  Friend's web server

3-2. How many leading octets must match for the “same network”?

The diagram above gives us the following relationships:

• Only the 1st octet matches (e.g., 192.x.x.x): a very loose grouping (/8). The other side could be a remote internet host you’ve never heard of.
• The 1st and 2nd octets match (e.g., 192.168.x.x): a slightly tighter grouping (/16).
• The 1st through 3rd octets match (e.g., 192.168.1.x): the network unit most commonly used by home LANs and small offices (/24). Devices in your home generally fit into this single /24.
• All four octets match: the exact same host (the same machine).

So 192.168.1.10 and 192.168.1.11 match through the 3rd octet (1), which puts them in the same network at the /24 level (a typical home LAN). On the other hand, 192.168.2.10 only matches up to the 2nd octet (168), so at /24 it’s on a different network — you’d need to widen the view to /16 before they’re “in the same network.”

3-3. What “being on the same network” actually lets you do

“On the same network” isn’t just about addresses looking similar — it has practical consequences:

• Communication within the same network is fast: packets reach the other party directly without going through a router, so response times are short.
• You can share printers and NAS: the reason your home printer or NAS (file server) is reachable instantly from any 192.168.1.x device is precisely because they’re on the same network.
• Reaching another network requires a router: any time you talk to a host on a different network, packets have to be forwarded through a router.

4. Network Part and Host Part — An IP Address Has Two Halves

In §3 we built up the intuition that “matching leading octets = same network.” In fact, IP addresses are designed to make this idea explicit, by splitting the address into two parts: the network part and the host part.

4-1. Network part (left) and host part (right)

Let’s split an IP address as follows.

For the same 192.168.1.10, the boundary (│) shifts depending on a mechanism called "CIDR" (covered later):

  With /8:   192 │ 168.1.10        Network part = only the 1st octet
  With /16:  192.168 │ 1.10        Network part = first 2 octets
  With /24:  192.168.1 │ 10        Network part = first 3 octets (typical home LAN)
  With /32:  192.168.1.10 │        All 32 bits are network = a single host

  Left side = Network part (shared address)  /  Right side = Host part (individual number)

• Network part: indicates which network this address belongs to. All devices on the same network share this part.
• Host part: identifies a specific device (host) within that network. Must be unique within the network.

It’s like an apartment building’s address plus an apartment number — the building’s address is the network part, and the apartment number is the host part. Everyone in the building shares the same building address; only the apartment numbers differ. That’s exactly what IP addresses do.

4-2. CIDR notation /24 — “the top N bits are the network part”

So where exactly is the boundary between network part and host part? That’s what CIDR notation (pronounced “cider”) makes explicit.

In CIDR notation, you append a slash and a number to the IP address (e.g., /24) to declare “the top N bits are the network part.”

192.168.1.10/24
            └─ Top 24 bits (= first 3 octets) are the network part
               Remaining 8 bits (= 4th octet) are the host part

The CIDR number is the number of bits in the network part. For /24, that’s 24 bits, so 24 ÷ 8 = 3 octets are network. Almost every home Wi-Fi router runs on /24, so for most practical purposes you can mentally equate “/24 = home LAN size.”

4-3. /8, /16, /24, /32 mapped to “matching octet count”

Notice how host count is inversely correlated with “network part bits”? The bigger the network part, the fewer bits remain for the host part, so the smaller the network capacity. /24‘s 256 hosts is the sweet spot for home use — “plenty of room for every PC, smartphone, and smart-home device under one roof” — which is why it’s the default for home routers.

4-4. Bonus: in-between boundaries like /20

So far we’ve only shown /8, /16, /24, and /32 — multiples of 8. Those line up nicely with octet boundaries, which makes them easy to read. But CIDR can actually split anywhere, not just at multiples of 8.

For example, /20 means “the top 20 bits are the network part” — only the first 4 bits of the 3rd octet belong to the network part, and the remaining 4 bits of that octet join the host part. This lets you design networks that sit between /16 (65,536 hosts) and /24 (256 hosts) — for instance, /20 gives you 4,096 hosts.

For beginners, the rule of thumb is fine: “home LAN is /24; for anything else, look it up when you encounter it.” The full mechanics of non-octet boundaries are something you’ll dig into properly only when you start working in network administration.

5. Private IP and Global (Public) IP

By now, a sharp question may have occurred to you: “My phone at home is 192.168.1.10, but my friend’s phone might be 192.168.1.10 too. If IP addresses are supposed to be unique worldwide, how is that OK?” Good catch — let’s resolve that here.

5-1. The 192.168.x.x on your home Wi-Fi — that’s a “private IP”

Addresses like 192.168.x.x, used inside home LANs, are special — they’re called private IPs, and they’re only valid inside your own home. By postal analogy, they’re like apartment numbers inside a single building: unique within the building, but only useful as a complete address when combined with the building’s street address.

By contrast, the worldwide-unique “real address on the internet” is called a global IP (or public IP). When you visit a website, the server you’re talking to always has a global IP.

5-2. The three reserved private IP ranges

Three address ranges are reserved worldwide as “private IP” — addresses you’re free to use inside your own network.

Home routers normally pick a /24 out of 192.168.0.0/16 for their LAN. The fact that your phone might be 192.168.1.10 while your neighbor’s phone is also 192.168.1.10 is no problem at all — within the private range, each home assigns its own internal addresses, and uniqueness is only required within each home.

5-3. Routers and NAT — the bridge between private and global

So how can a phone with only a private IP visit websites out on the global internet? The key is the router and a mechanism it runs behind the scenes called NAT (Network Address Translation).

[Inside your home]                    [The Internet]

Phone   192.168.1.10  ──┐
PC      192.168.1.11  ──┼─→ Router ──→ Global IP from your ISP
Printer 192.168.1.50  ──┘  (NAT)        (e.g., 203.0.113.5)
                                              │
                                              ↓
                                        Web server (e.g., 92.205.10.50)

When a device in your home reaches out to the internet, the router rewrites the source private IP to its own global IP before sending the packet out. When the reply comes back, the router consults a table it kept and rewrites the destination back to the appropriate private IP, so the response reaches the right device.

Thanks to this, every device in your home can share a single global IP to use the internet. NAT was originally a stopgap for “IPv4 address exhaustion,” but it also brought a side benefit: devices inside your home are not directly visible from the outside, which improves privacy.

6. How to Look Up Your Own IP Address

Now that you’ve got the theory, let’s actually look at your own IP addresses. Seeing both your private IP and global IP at the same time is the moment everything we’ve covered so far really clicks.

6-1. Finding your global IP

The easiest way is to open a browser and search for “my ip” or “what is my ip.” At the top of the search results, you’ll see the global IP you’re currently using to reach the internet (the address your router gets from your ISP).

If you’d rather use the command line, you can run the following from a terminal:

# Mac / Linux
curl ifconfig.me

# Windows (PowerShell)
Invoke-RestMethod ifconfig.me

You should see one line returned, like 203.0.113.5. That’s your home router’s global IP. If someone else in the house runs the same command on a different phone, they’ll get the same result — because every device in your home shares the same global IP when it goes out to the internet (as we explained in §5 with NAT).

6-2. Finding your private IP

The private IP assigned to you inside your home LAN is found differently on each operating system.

# Windows (Command Prompt or PowerShell)
ipconfig

# Mac / Linux (modern command)
ip addr show

# Mac / Linux (older command, still works)
ifconfig

The output lists multiple network interfaces. If you’re on Wi-Fi, look for entries like “Wi-Fi,” “en0,” or “wlan0.” For wired Ethernet, look for “Ethernet” or “eth0.” You should see a private IP like 192.168.x.x listed there.

Once you actually see a number like 192.168.1.10 on your own machine, the rest of this article should feel concrete — all of it is happening inside your home right now. If a family member runs the same command on their phone, you’ll likely see the same first three octets (e.g., 192.168.1.x) with only the last octet differing (e.g., .11 or .12). That’s exactly what “on the same /24 network” looks like.

7. IPv4 and IPv6 — Why a New Version Was Needed

So far we’ve focused on IPv4 addresses, but there’s a newer standard called IPv6. Let’s quickly cover why a new standard became necessary and how it differs from IPv4.

7-1. IPv4 address exhaustion — 4.3 billion wasn’t enough

As we mentioned in §1, an IPv4 address is 32 bits — that’s a total of about 4.3 billion possible addresses. Back in the 1980s, people figured “humanity will never run out.” Then the explosive growth of the internet — plus the era of multiple devices per person (PCs, smartphones, smart-home gadgets, IoT devices) — made 4.3 billion utterly inadequate.

New IPv4 allocations were exhausted region by region throughout the 2010s, and today you essentially can’t get fresh IPv4 addresses anymore. The stopgap that emerged in response is exactly the NAT and private IP setup we covered in §5. The reason a single home, with many devices, only needs one global IP is rooted in this exhaustion problem.

But a stopgap can’t solve the root cause forever — which is why IPv6, with a fundamentally larger address space, was introduced.

7-2. What IPv6 looks like and how groups work

IPv6 addresses are 128 bits long (four times IPv4) and look quite different.

IPv4:   192.168.1.10
        └─ 32 bits, four decimal numbers, dot-separated

IPv6:   2001:0db8:85a3:0000:0000:8a2e:0370:7334
        └─ 128 bits, eight hex groups, colon-separated

An IPv6 address is written as “eight groups of four-digit hexadecimal numbers,” with groups separated by colons (:) instead of dots. Each group is 16 bits, so where IPv4 had four 8-bit octets, IPv6 has eight 16-bit groups.

The total number of addresses in 128 bits is roughly 3.4 × 10³⁸. That’s “enough for 600 trillion addresses per square millimeter of Earth’s surface” — astronomical, with no realistic risk of running out.

7-3. The “grouping” idea is identical in IPv6

The format may have changed, but the principle is the same: matching leading bits = same network.

2001:0db8:85a3::/48      ← /48 is a typical home / small-network size in IPv6
└── 2001:0db8:85a3:0001::/64
    ├── 2001:0db8:85a3:0001::1     PC-A
    └── 2001:0db8:85a3:0001::2     PC-B

In IPv6, home networks commonly use /64 (so the top 64 bits are shared), and ISPs typically allocate /48 blocks to subscribers. The intuition you built up in §3 — “the more leading bits in common, the closer the network” — carries over directly.

7-4. IPv4 and IPv6 will coexist for the foreseeable future

“Will IPv4 just be retired then?” — the answer for now is no, not soon. The vast majority of devices, servers, and routers in the world still rely primarily on IPv4, and a hard cutover isn’t realistic.

The standard practice today is “dual stack” operation, where each device speaks both IPv4 and IPv6 and uses whichever is appropriate. Your phone, on your home Wi-Fi, very likely has both an IPv4 private IP (192.168.1.x) and an IPv6 global IP assigned to it.

8. Common Questions and Misconceptions

Finally, let’s clear up four points that beginners often stumble on.

8-1. Is my IP fixed, or does it change?

For home connections, in most cases it’s dynamic (it can change).

• Private IP (inside your home): assigned on the spot by the router via DHCP (automatic IP distribution), so it can change after a reboot. That said, with only a few devices on the LAN, the same IP usually keeps getting assigned in practice.
• Global IP (outward-facing): how often it changes depends on the ISP. On most US residential connections, it can stay the same for weeks or months unless the modem is unplugged for an extended time. Business contracts and “static IP” add-ons let you keep an IP that doesn’t change.

8-2. Can someone identify me from my IP?

The short answer: pinpointing a specific person is essentially impossible, but narrowing down to a region or ISP is feasible.

From a global IP, you can roughly tell things like “this IP belongs to a Comcast user in the New York area.” But identifying “John Smith on Maple Street” requires a formal disclosure request to the ISP (e.g., a court order) and isn’t something an ordinary person can do. On the other hand, website operators can see the IPs of visitors, so it’s safe to assume they have visibility into your approximate region and ISP.

8-3. Why can the same IP run multiple services? — port numbers

https://toolcluster.app/ is identified by an IP address plus a port number — two numbers together pin down a service.

• If the IP address is the building’s address,
• the port number is an “apartment number” inside the building.

On a single server (one IP), a website (port 443), email (port 25), SSH (port 22), and so on each live in their own numbered “apartment” — that’s how one IP can host many services at once. Detailed coverage of port numbers is outside the scope of this article and worth its own piece.

8-4. What about “IP classes (Class A/B/C)” — I’ve heard of those?

In older IPv4, addresses were mechanically grouped into Class A (/8), Class B (/16), and Class C (/24) based on the value of the 1st octet. But after CIDR (the notation we covered in §4) was introduced in 1993, this old class system is effectively no longer used. It’s now a “classful” legacy concept and you don’t really need to learn it. If you spot it in old documentation, just read it as “ah, this is pre-CIDR talk” and move on.

Summary — Four-Line Essence of IP Addressing

This was a long article, but its essence boils down to four lines:

• An IPv4 address is 32 bits split into 8-bit × 4 octets — an “address on the internet.”
• Bigger groups on the left, individual hosts on the right — share the leading N octets and you’re on the same network.
• CIDR /24 means “the top 24 bits = 3 octets are the network part,” and is the standard for home LANs.
• Inside the home you have a private IP; outward-facing you have a global IP — the router bridges them with NAT.

Keep these in mind, and the next time you open your router admin page, network settings, or a server log, what looked like “a meaningless string of numbers” should now feel readable.

From here, the world of networking branches further into DNS (the name-to-IP mapping), ports, firewalls, VPNs, routing, and more — but the IP address fundamentals are the foundation under all of them. The “group structure” intuition you picked up here will keep paying off.

For the broader topic of how data is represented numerically, see also How to Choose SQL Numeric Types Correctly on this site. The question of “how many bits to use for an integer” is closely related to IP addressing.

FAQ

Q1. Is there a way to change my IP address?

A. For your global IP, powering your home router off for at least 5–10 minutes and then back on often results in a new IP (the ISP’s DHCP lease expires and a fresh one is assigned). For private IPs, you can change them by reserving or excluding specific devices in your router’s DHCP settings. With a VPN, you can effectively swap the global IP that the outside world sees for the VPN provider’s IP.

Q2. Is IPv6 ready for everyday use?

A. Major ISPs (Comcast, Verizon, AT&T, Spectrum, etc.) provide IPv6 as standard, and if your router supports it, IPv6 traffic happens automatically. Major services like Google, YouTube, and Facebook are all IPv6-ready, so plenty of users are already using IPv6 without realizing it. If you see an inet6 line in ip addr show, you’re already on IPv6.

Q3. How can I see all the devices on my network?

A. Log into your router’s admin page (often http://192.168.1.1/ or http://192.168.0.1/) and look for sections like “Connected devices” or “DHCP client list.” There you’ll see the IP and MAC addresses of every device currently on your home network. From a command line, arp -a gives a similar view.

Q4. How many bytes is an IPv4 address?

A. 4 bytes (= 32 bits). One octet equals one byte, so four octets is 4 bytes total. When storing an IP in a database, you can keep it as a string (VARCHAR(15)), but for space efficiency it’s common to store it as a 32-bit integer (INT UNSIGNED). The same line of thinking ties into How to Choose SQL Numeric Types Correctly.

Q5. What happens to my IP when I use a VPN?

A. With a VPN (Virtual Private Network), your traffic goes out to the internet via the VPN server, so the source IP that other parties see is the VPN server’s global IP. If you’re located in the US but connect to a VPN server in Germany, German-region websites will see you as coming “from Germany.” It’s a standard technique for temporarily hiding your home’s global IP, but the VPN provider itself can potentially see your traffic, so choosing a trustworthy provider matters.