I am following the code in this page: https://www.twilio.com/en-us/blog/asynchronous-http-requests-in-python-with-aiohttp.
Here is the first approach using aiohttp:

import aiohttp
import asyncio
import time

start_time = time.time()


async def main():
    async with aiohttp.ClientSession() as session:
        for num in range(1, 10):
            pokemon_url = f'https://pokeapi.co/api/v2/pokemon/{num}'
            async with session.get(pokemon_url) as response:
                pokemon_json = await response.json()
                print(pokemon_json['name'], num)


asyncio.run(main())

print(f'{time.time() - start_time} seconds')

If I understand it correctly, the for loop executes session.get synchronously thus the GET request are sent out sequentially. Asynchronous Programming is not giving ertra performance benefits here.

Approach 2 is the regular approach using requests. It’s also executed synchronously:

import requests
import time

start_time = time.time()

for num in range(1, 10):
    url = f'https://pokeapi.co/api/v2/pokemon/{num}'
    response = requests.get(url)
    pokemon = response.json()
    print(pokemon['name'])

print(f'{time.time() - start_time} seconds')

The approach 1 took 3.581976890563965 seconds. Approach 2 took 8.428680419921875 seconds. Since they both executes GET requests sequentially, what’s the reason for approach 1
using aiohttp still being faster than requests?