Day 1: Node.js + Claude API: Your First AI-Powered App | Build AI Apps with JavaScript

Why JavaScript for AI Apps

JavaScript is the language of the web. If you want to build AI-powered web apps — chatbots, writing tools, document processors — JavaScript is the fastest path from idea to deployed product. The Claude API has a first-class Node.js SDK, and the same code runs on the frontend and backend.

Setup

Terminal

mkdir ai-js-app && cd ai-js-app
npm init -y
npm install @anthropic-ai/sdk dotenv
touch .env index.js

.env

ANTHROPIC_API_KEY=sk-ant-your-key-here

index.js — First API Call

import Anthropic from '@anthropic-ai/sdk';
import * as dotenv from 'dotenv';
dotenv.config();

const client = new Anthropic();

async function chat(userMessage) {
  const message = await client.messages.create({
    model: 'claude-opus-4-5',
    max_tokens: 1024,
    messages: [
      { role: 'user', content: userMessage }
    ]
  });
  
  return message.content[0].text;
}

// Test it
const reply = await chat('Explain async/await in JavaScript in 3 sentences.');
console.log(reply);

package.json — Add type: module

{
  "type": "module",
  "scripts": {
    "start": "node index.js"
  }
}

Run It

node index.js

Understanding the Response Object

Inspect the Full Response

const message = await client.messages.create({ ... });

console.log(message.id);              // msg_xxx
console.log(message.model);           // claude-opus-4-5
console.log(message.stop_reason);     // end_turn
console.log(message.usage.input_tokens);   // tokens used
console.log(message.usage.output_tokens);  // tokens generated
console.log(message.content[0].type);      // 'text'
console.log(message.content[0].text);      // the actual reply

Day 1 Exercise

Build a Simple CLI Question Answerer

Set up the project and create index.js with the code above.
Run it and verify you get a response.
Modify the question and run it 3 more times.
Log the full message object and read through the structure.
Build a simple loop that reads questions from command line args and prints responses.

Day 1 Summary

Anthropic SDK installed and configured with .env for the API key.
client.messages.create() is the core method — model, max_tokens, messages array.
Response contains content array, usage stats, and stop_reason.
Use ES modules (type: module in package.json) for clean import syntax.

Day 2: Streaming Responses

→