Velamuvi AIAcademy

Lesson 2 of 12 · How the Web Talks

HTTP Requests and Responses

See what an HTTP request and response really look like: URLs, the request line, status line, headers and body, and why HTTP is stateless. With runnable Python examples.

  • 8 min read
  • Beginner
  • Reviewed
In this lesson you will learn
  • What HTTP and HTTPS are
  • The parts of a URL
  • What is inside an HTTP request and an HTTP response
  • How to pull a response apart in Python
  • Why HTTP is called stateless, and why that matters for AI chat

The language of the web

In the last lesson, your program sent a request and got a response. But how do two programs that have never met understand each other? They both speak the same language: HTTP.

Every web page you open and every API you call uses HTTP. Always use the https:// address of an API, especially when your requests contain passwords or keys.

The parts of a URL

A URL has several parts, and Python can split them for you:

Try it yourselfPython
from urllib.parse import urlparse

url = "https://api.open-meteo.com/v1/forecast?latitude=17.38&longitude=78.49"
parts = urlparse(url)

print("Scheme:", parts.scheme)
print("Host:  ", parts.netloc)
print("Path:  ", parts.path)
print("Query: ", parts.query)
PartExampleMeaning
Schemehttpswhich rules to use: HTTP or secure HTTPS
Hostapi.open-meteo.comwhich computer (server) to talk to
Path/v1/forecastwhich endpoint on that server
Querylatitude=17.38&longitude=78.49extra details for this request

You will learn all about the query part in the Query Parameters lesson.

What a request looks like

An HTTP request is just text, written in a fixed layout. When your weather program ran, it sent something like this:

GET /v1/forecast?latitude=17.38&longitude=78.49&current=temperature_2m HTTP/1.1
Host: api.open-meteo.com
Accept: application/json

It has up to four parts:

  1. The request line: the method (GET), the path and the HTTP version.
  2. Headers: extra information, one per line, written as Name: value.
  3. A blank line, which means "the headers are finished".
  4. A body with data to send. A GET request usually has no body.

What a response looks like

The server answers in the same kind of layout:

HTTP/1.1 200 OK
Content-Type: application/json
Date: Tue, 17 Nov 2026 11:05:00 GMT

{"latitude": 17.39, "longitude": 78.46, "current": {"time": "2026-11-17T16:30", "temperature_2m": 28.5}}

Then come the headers, a blank line and the body with the actual data.

Taking a response apart in Python

Because a response is text in a fixed layout, you can take it apart yourself. This is exactly what libraries such as requests do for you behind the scenes.

Try it yourselfPython
import json

raw = """HTTP/1.1 200 OK
Content-Type: application/json
Date: Tue, 17 Nov 2026 11:05:00 GMT

{"city": "Hyderabad", "temperature": 28.5}"""

head, body = raw.split("\n\n", 1)        # the blank line separates them
lines = head.split("\n")

status_line = lines[0]
print("Status line:", status_line)
print("Status code:", status_line.split(" ")[1])

for line in lines[1:]:
    name, value = line.split(": ", 1)
    print("Header:", name, "=", value)

data = json.loads(body)                  # turn JSON text into a dictionary
print("Temperature:", data["temperature"])

json.loads() turns JSON text into a Python dictionary. You will learn JSON properly in its own lesson.

Seeing a real response

Here is the same idea with a real server. The requests library has already split the response into its parts for you:

Try it yourselfPython
import requests

response = requests.get(
    "https://api.open-meteo.com/v1/forecast",
    params={"latitude": 17.38, "longitude": 78.49, "current": "temperature_2m"},
    timeout=10,
)

print("Status code: ", response.status_code)
print("Content-Type:", response.headers["Content-Type"])
print("Body:        ", response.text[:120])

HTTP is stateless

Think of a shop assistant with no memory at all. Every time you walk up, you must say who you are and what you want, even if you were there a minute ago.

This has a big effect on AI chat apps. When you chat with a model through its API, the server does not remember the conversation. Your program sends the whole conversation so far with every new message. That is why long chats cost more: every request carries all the earlier messages again.

Common mistakes

Practice exercise

Take an error response apart, just like an HTTP library does.

Your solutionPython
import json

raw = """HTTP/1.1 404 Not Found
Content-Type: application/json

{"error": "City not found", "city": "Atlantis"}"""

# 1. Split raw into the head and the body. They are separated by a blank line.
# 2. From the status line, print the status code as a number and the reason phrase.
# 3. Turn the body into a dictionary and print the error message.

# Output must be exactly:
# Status code: 404
# Reason: Not Found
# Error: City not found
Knowledge check

What separates the headers from the body in an HTTP message?

Summary

  • HTTP is the set of rules for writing requests and responses. HTTPS is the secure, encrypted version.
  • A URL has a scheme, a host, a path and sometimes a query.
  • A request has a request line, headers, a blank line and an optional body.
  • A response has a status line, headers, a blank line and a body.
  • HTTP is stateless: every request must carry everything the server needs.

Next: the different kinds of request, from reading data to deleting it.

Finished reading? Tick it off to track your progress.