How-To Guide

OpenAI API Setup

Every AI app starts with an API key. Here's how to get one and make your first request in 10 minutes.

📋 What You'll Need

Node.js 18+ or Python 3.8+ — see our setup guides
A credit card (OpenAI is paid, but very cheap to start)
10 minutes

Cost warning: The OpenAI API is NOT free. You pay per token. But most hobby projects cost a few dollars per month. Load $5 to start — that's enough for hundreds of chats.

Create an OpenAI Account

Go to the platform:

👉 platform.openai.com

Add Billing

Go to Settings → Billing. Add a payment method and load a small balance ($5 is plenty).

Set a monthly usage limit so you never get surprised. $10/month is safe for learning.

Create an API Key

Go to API Keys → Create new secret key. Give it a name like "My First App".

Copy the key immediately. It looks like sk-... — you'll never see it again.

💡 Store it in .env, not in code. Never commit API keys to GitHub.

Install the SDK

Node.js:

npm install openai dotenv

Python:

pip install openai python-dotenv

Make Your First Call (Node.js)

Create .env:

OPENAI_API_KEY=sk-your-key-here

Create chat.js:

import 'dotenv/config';
import OpenAI from 'openai';

const openai = new OpenAI({
    apiKey: process.env.OPENAI_API_KEY
});

const response = await openai.chat.completions.create({
    model: 'gpt-4o-mini',
    messages: [
        { role: 'user', content: 'Explain recursion in one sentence.' }
    ]
});

console.log(response.choices[0].message.content);

Run it:

node chat.js

You should see the AI's reply. 🎉

🧠 Essential Concepts

Models — gpt-4o-mini is cheap and fast. gpt-4o is more capable. gpt-4-turbo is middle ground.
Messages — an array of {role, content} objects. Roles: system, user, assistant.
Tokens — roughly 0.75 words. You pay per input + output token.
Temperature — 0 = deterministic, 1 = creative, 2 = chaotic.
System prompt — sets the AI's behavior ("You are a helpful coding tutor").

💬 System Prompts Matter

const response = await openai.chat.completions.create({
    model: 'gpt-4o-mini',
    messages: [
        {
            role: 'system',
            content: 'You are a patient programming tutor. Always reply in 2 sentences or less.'
        },
        { role: 'user', content: 'What is a promise?' }
    ],
    temperature: 0.7
});

The system message shapes every reply. This is how you build "chatbots" with personality.