AI APIs Cheat Sheet

Your complete quick reference for working with AI APIs. Bookmark this page!

Side-by-Side API Comparison

Installation

# Install all three
npm install openai @anthropic-ai/sdk @google/generative-ai

# Or install individually
npm install openai
npm install @anthropic-ai/sdk
npm install @google/generative-ai

Environment Setup

# .env
OPENAI_API_KEY=sk-proj-your-key-here
ANTHROPIC_API_KEY=sk-ant-your-key-here
GOOGLE_API_KEY=AIzaSy-your-key-here

Code Templates

OpenAI — Basic Chat

import OpenAI from "openai";

const openai = new OpenAI();

const response = await openai.chat.completions.create({
  model: "gpt-4o",
  messages: [
    { role: "system", content: "You are a helpful assistant." },
    { role: "user", content: "Hello!" },
  ],
});

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

OpenAI — Streaming

const stream = await openai.chat.completions.create({
  model: "gpt-4o",
  messages: [{ role: "user", content: "Hello!" }],
  stream: true,
});

for await (const chunk of stream) {
  const content = chunk.choices[0]?.delta?.content;
  if (content) process.stdout.write(content);
}

Anthropic — Basic Chat

import Anthropic from "@anthropic-ai/sdk";

const anthropic = new Anthropic();

const message = await anthropic.messages.create({
  model: "claude-sonnet-4-20250514",
  max_tokens: 1024,
  system: "You are a helpful assistant.",
  messages: [
    { role: "user", content: "Hello!" },
  ],
});

console.log(message.content[0].type === "text" ? message.content[0].text : "");

Anthropic — Streaming

const stream = anthropic.messages.stream({
  model: "claude-sonnet-4-20250514",
  max_tokens: 1024,
  messages: [{ role: "user", content: "Hello!" }],
});

stream.on("text", (text) => process.stdout.write(text));
await stream.finalMessage();

Google Gemini — Basic Chat

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI(process.env.GOOGLE_API_KEY!);
const model = genAI.getGenerativeModel({
  model: "gemini-2.5-flash",
  systemInstruction: "You are a helpful assistant.",
});

const result = await model.generateContent("Hello!");
console.log(result.response.text());

Google Gemini — Streaming

const result = await model.generateContentStream("Hello!");

for await (const chunk of result.stream) {
  process.stdout.write(chunk.text());
}

Google Gemini — Chat

const chat = model.startChat({ history: [] });

const result1 = await chat.sendMessage("Hello!");
console.log(result1.response.text());

const result2 = await chat.sendMessage("Tell me more.");
console.log(result2.response.text());

Pricing Comparison

Cost Tiers

Model Selection Guide

By Task

By Priority

Response Access Patterns

// OpenAI
const text = response.choices[0].message.content;
const tokens = response.usage?.total_tokens;

// Anthropic
const text = message.content[0].type === "text" ? message.content[0].text : "";
const tokens = message.usage.input_tokens + message.usage.output_tokens;

// Gemini
const text = result.response.text();
const tokens = result.response.usageMetadata?.totalTokenCount;

Error Handling Template

async function callAI(provider: "openai" | "anthropic" | "gemini") {
  try {
    // Your API call here
  } catch (error: any) {
    if (error?.status === 429 || error?.message?.includes("429")) {
      console.error("Rate limited — implement retry with backoff");
    } else if (error?.status === 401 || error?.message?.includes("API_KEY")) {
      console.error("Authentication failed — check your API key");
    } else if (error?.status === 400) {
      console.error("Bad request — check your parameters");
    } else if (error?.status >= 500) {
      console.error("Server error — retry later");
    } else {
      throw error;
    }
  }
}

Retry with Exponential Backoff

async function withRetry<T>(
  fn: () => Promise<T>,
  maxRetries = 3,
  baseDelay = 1000
): Promise<T> {
  for (let i = 0; i <= maxRetries; i++) {
    try {
      return await fn();
    } catch (error: any) {
      if (error?.status === 429 && i < maxRetries) {
        const delay = baseDelay * Math.pow(2, i);
        await new Promise((r) => setTimeout(r, delay));
      } else {
        throw error;
      }
    }
  }
  throw new Error("Unreachable");
}

---

AI Prompts for API Integration

Getting Started

• "Set up a Node.js project with TypeScript that can call the [OpenAI/Anthropic/Google] API. Include environment variables and error handling."
• "I want to compare responses from OpenAI, Anthropic, and Google for the same prompt. Build a script that calls all three and shows the results."
• "Create a reusable AI client class that supports multiple providers. I should be able to switch between OpenAI, Anthropic, and Google easily."

Building Features

• "Build a chat API endpoint that supports streaming responses using [provider]. The frontend is built with React."
• "Create a text summarization feature using [provider]. It should accept long text and return a 3-sentence summary."
• "Build a code review tool that analyzes code and returns structured feedback as JSON using OpenAI's structured output."
• "Create a function-calling setup with OpenAI where the AI can search a database and return results."

Production Readiness

• "Add rate limiting, retry logic, and error handling to my AI API calls. I'm using [provider]."
• "Set up cost tracking for my AI API usage. Log tokens used per request and estimate daily costs."
• "Build a prompt caching layer that stores AI responses in [Redis/Firestore] to avoid redundant API calls."
• "Create a fallback system that tries OpenAI first, then falls back to Anthropic if it fails."

Optimization

• "My AI API costs are too high. Here's my current setup: [describe]. How can I reduce costs?"
• "Optimize my prompts for token efficiency. Here are my current prompts: [paste]. Make them shorter without losing quality."
• "I'm hitting rate limits. Help me implement request queuing with concurrency control."

Debugging

• "I'm getting a 429 error from [provider]. What does this mean and how do I fix it?"
• "My streaming implementation isn't working. Here's my code: [paste]. What's wrong?"
• "The AI response doesn't match my expected JSON format. Here's my prompt and the response: [paste]. How do I fix this?"
• "My API key works in curl but not in my Node.js app. Here's my code: [paste]."

Quick Start Checklist

1. SIGN UP for an API account (OpenAI, Anthropic, or Google)
2. GET your API key from the provider's dashboard
3. INSTALL the SDK: npm install [package]
4. CREATE a .env file with your API key
5. ADD .env to .gitignore
6. SET a spending limit on your account
7. WRITE your first API call
8. TEST with a cheap model first (mini/haiku/flash)
9. ADD error handling and retries
10. UPGRADE to a better model if needed

Key URLs

---

You now have everything you need to integrate AI into your applications. Start with one provider, build something small, and expand from there. The best way to learn AI APIs is to build with them.

Happy building!

---

🌐 www.genai-mentor.ai