Nexus Learn

Log in · Create free account

Module 6: Calling an AI service from Python · 11 min

Sending a prompt to a model

Every AI company has its own address and exact message format, so always copy the current shape from the provider's documentation. The pattern is the same everywhere: send your prompt and key as JSON, read the JSON reply, and handle errors.

import os
import requests

def ask(prompt):
    url = "https://api.example.com/v1/chat"   # use your provider's real address
    headers = {"Authorization": "Bearer " + os.environ["MY_AI_KEY"]}
    body = {"model": "model-name", "messages": [{"role": "user", "content": prompt}]}
    r = requests.post(url, headers=headers, json=body, timeout=60)
    r.raise_for_status()      # stop with a clear error if status is not OK
    return r.json()           # then pick the answer text out of this data

# print(ask("Explain a Python list in one sentence."))

Many providers also offer an official Python package that does these steps for you. Combine this lesson with the Prompt Engineering and Agentic AI courses to build real tools.

You have now finished the lessons. Take the final exam to earn your certificate.

(free account needed to save progress)

← Keeping API keys safe