Internet vs Web
The internet is the global network of computers — the cables, routers, and protocols that let any two machines on it exchange data. The World Wide Web (the "web") is one application that runs on the internet: a system of linked pages and resources viewed in a browser. The internet was decades old before the web existed; even today, plenty of internet traffic isn't web traffic at all.
Last reviewed on 2026-04-27.
Quick Comparison
| Aspect | Internet | Web |
|---|---|---|
| What it is | A worldwide network of interconnected networks | An information system of linked pages running on the internet |
| Layer | Infrastructure: hardware, links, addressing, routing | Application: documents, links, browsers, servers |
| Primary protocols | IP, TCP, UDP, DNS | HTTP and HTTPS |
| Started | ARPANET in 1969; "the internet" emerged through the 1980s | Proposed in 1989; first browser in 1990 |
| Examples of use | Web browsing, email, video calls, online games, streaming, file transfer | Reading a website in a browser, web apps, REST APIs over HTTPS |
| Can exist without the other? | Yes — and did, for decades before the web | No — the web requires the internet to run on |
Key Differences
1. Network versus application
The internet is plumbing. It's the worldwide system of routers and links and the agreements (protocols) by which any two computers on the network can identify each other and exchange packets of data. IPv4 and IPv6 are how machines find each other; TCP and UDP are how they hold conversations.
The web is one of many applications that uses that plumbing. It's a system of documents (web pages) addressed by URLs, served by web servers, and read by browsers using the HTTP and HTTPS protocols. The web is enormous and obvious to most users, but it sits on top of the internet, not instead of it.
2. Different histories
The internet traces back to ARPANET, a U.S. research network that exchanged its first message in 1969. Through the 1970s and 1980s the protocols that became TCP/IP were developed, more networks connected, and the public, decentralised "internet" emerged.
The web arrived much later. Tim Berners-Lee proposed it at CERN in 1989 and built the first browser, server, and HTML in 1990. By then the internet had already been moving email, files, and Usenet posts for two decades. The web added a friendly hyperlinked layer that made the underlying network usable for non-technical people.
3. Internet things that aren't the web
This is where the distinction stops being academic. All of the following use the internet but are not part of the web:
- Email (SMTP, IMAP, POP3).
- Video calls (WebRTC media streams, SIP).
- Online games using their own UDP-based protocols.
- SSH and remote terminal sessions.
- BitTorrent and other peer-to-peer file sharing.
- DNS itself — the lookup that turns
diffbetween.netinto an IP address. - Most "internet of things" telemetry between devices and clouds.
Open a chat app, a video call, or a multiplayer game and you're using the internet, but the web (in the strict sense) is barely involved.
4. URLs, browsers, and HTTP
What identifies the web as the web is the combination of three pieces:
- URLs (Uniform Resource Locators) — addresses for resources, like
https://diffbetween.net/technology/internet-vs-web/. - HTTP/HTTPS — the protocol used to request and receive those resources. See HTTP vs HTTPS for what HTTPS adds.
- HTML — the document format that lets pages link to each other, which is what makes it a "web."
If you can describe a thing in a URL and fetch it with HTTP, it's part of the web. If you reach it some other way (a TCP socket, a custom protocol, a peer-to-peer swarm), it's on the internet but not on the web.
5. Why the distinction still matters
It matters whenever someone says "the internet is down." Often what's down is the web (a particular site won't load) or DNS (your computer can't find the site). The actual underlying network is usually fine. Knowing which layer is broken is the first step to fixing anything.
It also matters when reading news or policy. Discussions about "regulating the internet" usually turn out to mean regulating particular web services. The internet itself — a decentralised system of networks — is much harder to regulate as a single thing than a set of named platforms running on top of it.
Worked Example: Loading a Web Page
Type https://diffbetween.net into a browser and press Enter. Here's the rough order of events, with internet-vs-web layers labelled:
- DNS lookup (internet): your computer asks a DNS server for the IP address of
diffbetween.net. - TCP/QUIC handshake (internet): your browser opens a connection to that IP address, using TCP or UDP as the transport.
- TLS handshake (internet): the browser and server agree on encryption keys for HTTPS.
- HTTP request (web): the browser asks the server for
/using HTTPS. - HTML response (web): the server returns an HTML document, which the browser parses and renders.
- Linked resources (web): the browser follows links inside the HTML to fetch CSS, JavaScript, images — each its own HTTP request.
Steps 1–3 are pure internet machinery. Steps 4–6 are the web. Strip away the browser and HTML and the internet still works; strip away the internet and there's no web at all.
Common Misconceptions
- "The web and the internet are the same thing." They're not. The web is the most visible application on the internet, but it's not the only one — and it's much younger than the network underneath it.
- "The dark web is a different internet." It's not. Tor and similar anonymity networks run on the same internet as everything else; they just route traffic through additional layers to obscure who's talking to whom.
- "Apps don't use the web." Most mobile and desktop apps make HTTP/HTTPS calls to APIs behind the scenes, so a lot of "non-web" experiences are technically built on the web's protocols even though there's no browser involved.
- "Web 3.0 means a new internet." Most uses of the term refer to new applications running on the existing internet, not a replacement for the underlying network.