🦁 IP Animals
🔌 Networking Fundamentals

The OSI Model Explained in Plain English

The OSI model is a seven-layer map of how data moves across a network, from the electrical signals on a wire up to the app on your screen. Here is each layer explained in plain English, with real examples you already use every day.

The OSI model, short for Open Systems Interconnection model, is a way of breaking networking into seven tidy layers. Instead of treating "the internet" as one incomprehensible blob, it divides the work of moving data into distinct stages, each with a single job. This layered thinking is one of the most useful mental tools in all of networking, and once it clicks, topics like HTTP, routing, and Wi-Fi all fall into place.

Why layers exist at all

Imagine posting a letter. You write the words, put them in an envelope, add an address, hand it to the post office, and it travels by van, plane, and foot before someone opens it and reads your words. Each of those steps is handled by someone who does not need to understand the others. The van driver does not read your letter; you do not need to know the flight schedule.

Networking works the same way. Each layer of the OSI model provides a service to the layer above it and relies on the layer below. This separation means engineers can improve or replace one layer — say, swapping a copper cable for fibre — without rewriting everything else.

Key fact

The OSI model is a reference, not the software the internet actually runs. The real internet uses the simpler TCP/IP model. But OSI's seven-layer vocabulary is so useful that engineers everywhere describe issues as being at "layer 2" or "layer 7" every single day.

The seven layers at a glance

The layers are numbered from the bottom up. Layer 1 is closest to the physical hardware; Layer 7 is closest to you, the user. Data created by an app travels down the stack on the sending device and back up the stack on the receiving device.

#LayerJobReal-world example
7ApplicationThe interface apps use to reach the networkHTTP, DNS, email (SMTP)
6PresentationFormatting, encryption, compressionTLS encryption, JPEG, UTF-8
5SessionOpening, managing and closing conversationsKeeping a login session alive
4TransportReliable or fast end-to-end deliveryTCP and UDP, port numbers
3NetworkLogical addressing and routing between networksIP addresses, routers
2Data LinkLocal delivery on a single linkMAC addresses, Ethernet, Wi-Fi
1PhysicalRaw bits as signals on a mediumCables, radio waves, fibre light

Layers 1 to 3: getting bits to the right place

Layer 1, Physical. This is the raw stuff: the electrical voltages, pulses of light, or radio waves that carry ones and zeros. Ethernet cables, fibre strands, and Wi-Fi radios all live here. If a cable is unplugged, that is a Layer 1 problem.

Layer 2, Data Link. This layer moves data between two directly connected devices on the same local network. It uses MAC addresses — the hardware identifiers burned into every network adapter — to deliver frames to the right device on your Wi-Fi or Ethernet segment. Switches operate here.

Layer 3, Network. This is the layer of the IP address. While Layer 2 handles the local hop, Layer 3 handles the big picture: getting a packet from your home network across the world to a distant server. Routers live here, forwarding packets based on their destination IP. This is exactly the process described in our guide to how data travels the internet.

Layer 4: reliable delivery from end to end

The Transport layer takes responsibility for the conversation between the two end devices as a whole. This is where TCP and UDP operate. TCP guarantees that every piece of data arrives, in order and intact, by numbering packets and re-sending anything lost. UDP trades those guarantees for raw speed, which suits live video and games.

Layer 4 also introduces port numbers, which let a single device juggle many conversations at once — your browser, email client, and music app can all share one IP address because each uses a different port.

Layers 5 to 7: making data meaningful

Layer 5, Session. This layer sets up, coordinates, and tears down the ongoing dialogue between two applications — think of it as managing the start and end of a "conversation," including keeping you logged in.

Layer 6, Presentation. Here data is translated into a form both sides understand: character encoding, compression, and crucially encryption. When your browser negotiates TLS/SSL to secure a connection, that work sits at this layer.

Layer 7, Application. The top layer is the one closest to you. It is not the app itself but the networking interface the app uses — protocols like HTTP for web pages, DNS for name lookups, and SMTP for email. When you type a web address and hit enter, you are handing a job to Layer 7.

How OSI maps to the real internet

The internet does not run OSI directly — it runs the TCP/IP model, which compresses these seven layers into about four practical ones. Roughly, TCP/IP's Application layer covers OSI layers 5 to 7, its Transport layer matches OSI Layer 4, its Internet layer matches OSI Layer 3, and its Link layer covers OSI layers 1 and 2.

So why learn OSI at all? Because it is the shared language of networking. When a network engineer says a firewall works at "layer 4" or a load balancer at "layer 7," they are pointing to a precise idea instantly understood across the whole industry. That common vocabulary is the OSI model's lasting gift, and it makes the rest of the topics here on IP Animals far easier to follow.

Frequently asked questions

What are the seven layers of the OSI model?

From bottom to top they are: Physical, Data Link, Network, Transport, Session, Presentation, and Application. Each layer handles one part of moving data across a network and passes its work to the layer above or below.

Is the OSI model actually used on the internet?

The internet runs on the TCP/IP model rather than OSI, but the OSI model is still widely taught and used as a common vocabulary. Engineers routinely describe problems as being at "layer 2" or "layer 7", and those numbers come straight from OSI.

What is the difference between the OSI model and TCP/IP?

The OSI model has seven layers and is a conceptual reference. TCP/IP is the practical model the internet actually uses, with four or five layers that combine several OSI layers. TCP/IP was built to solve real problems, while OSI was designed as a complete theoretical framework.

What layer does an IP address belong to?

IP addresses live at Layer 3, the Network layer. That is the layer responsible for logical addressing and routing packets between different networks so they can reach a destination anywhere in the world.

How do I remember the OSI layers?

A popular mnemonic from top to bottom is "All People Seem To Need Data Processing" — Application, Presentation, Session, Transport, Network, Data Link, Physical. Reversed, "Please Do Not Throw Sausage Pizza Away" runs from Physical up to Application.

Curious what your own IP is? Visit the IP zoo →