HTTP Request
An HTTP request is the request that your device sends to a remote computer on the internet to get something — a webpage, an image, or a response from a program. It contains exactly one address, a desired action, and additional details, and it is always followed by a response.
When you type in an internet address, nothing happens by magic. Your device sends a written request to another computer that holds the page ready. This request is called an HTTP request. HTTP is just the name of the language both sides have agreed on so they can understand each other. At its core, the request says: which address is meant, and what should happen with it. The other computer reads this, does something, and sends a response back. This pair of request and response is the basic process on which the entire web is built.
The smallest unit in which the web is billed
Almost everything you do online breaks down into such requests. A single news page often triggers 50 to 200 requests: one for the text, many for images, fonts, and ads. That’s why the number of requests is a key reason a page loads fast or slow. Each request costs time because it has to travel physically through cables.
For companies, requests are also a unit used to make money. Cloud service providers — that is, providers of rented computing power in remote data centers — often bill per million requests. AI providers work the same way: every question to a language model is technically an HTTP request to their servers. When a report states that a service processes billions of requests per day, this is exactly what is meant.
Requests are also the point where attacks take aim. In a denial-of-service attack, thousands of hijacked devices send pointless requests simultaneously until the server can no longer respond. The individual request is harmless — the sheer mass is the problem.
What’s actually inside a request
A request is basically a short text with a fixed structure. The first line contains the method, i.e. the desired action. GET means “send me something,” POST means “here’s data, take it.” Next to it is the exact address of the desired content. When submitting a form, it’s almost always a POST.
After that come the headers, a list of additional details. These state, for example, which language you prefer, which browser you are, and whether you’re logged in. Being logged in works via small snippets of text that the server previously gave you and that you send along with every request. The server itself doesn’t remember anything about you between two requests.
The response has the same structure and begins with a three-digit number. 200 means: everything’s fine. 404 means: the requested address doesn’t exist. 500 means: the server has an error of its own. A common misconception is that an HTTP request is a persistent connection. It’s more like a letter with a return receipt than a phone call.
Where you notice requests in everyday life
You notice them most clearly when something goes wrong. The error page “404 – Page not found” is the response to a request that went nowhere. The block screen for too many clicks in a short time is also a response, namely the code 429 for “too many requests.”
Every browser also has built-in developer tools, usually accessible via the F12 key. There, in the network tab, you can see every single request of a page along with its size and duration. That’s the fastest way to not just read about the term, but actually see it.
In news and product descriptions, the term often comes up in connection with interfaces, i.e. access points through which programs supply each other with data. When a provider states a limit of 60 requests per minute, they mean HTTP requests. Apps on your phone work the same way: they show you a nice interface, but fetch the content via the very same requests as a browser.
