Setup & First API Calls

Day 17 of 80

What You'll Accomplish Today

• Install the Anthropic SDK and python-dotenv
• Get your API key from Anthropic and store it safely in a .env file
• Make your first successful Claude API call from Python
• Inspect the full response object — not just the text
• Use system prompts to give Claude a persona
• Build dynamic prompts with f-strings and variables
• Handle API errors gracefully with try/except

Part 1: Install the SDK (5 min)

Open your terminal in Cursor (Ctrl + `) and run:

$ pip install anthropic python-dotenv

anthropic is the official Python SDK — it handles all the HTTP complexity for you. python-dotenv is a utility that reads a .env file and loads its values as environment variables, which is how you'll keep your API key out of your code.

Part 2: Get Your API Key (5 min)

1. Go to the Anthropic Console
   Visit console.anthropic.com. Create an account if you don't have one — free tier credits are provided automatically.
2. Generate an API key
   In the console, go to API Keys and click Create Key. Give it a name like "Python Training". Copy the key — it starts with sk-ant-.
3. Create your .env file
   In your project folder, create a new file called exactly .env (the dot is part of the name). Add this single line:
   .env

   ANTHROPIC_API_KEY=sk-ant-your-key-here

   Replace sk-ant-your-key-here with your actual key. No quotes needed.

Part 3: Five Cells in Jupyter

Open your terminal and launch Jupyter:

$ jupyter notebook

Create a new notebook called week-4-practice.ipynb. Work through each cell below — run it, read the output, then read the explanation before moving to the next.

Cell 1 — Your First API Call

import anthropic
from dotenv import load_dotenv

load_dotenv()
client = anthropic.Anthropic()

message = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=100,
    messages=[{
        "role": "user",
        "content": "In one sentence, what is an API?"
    }]
)

print(message.content[0].text)

Cell 2 — Inspect the Full Response

print(f"Model: {message.model}")
print(f"Stop reason: {message.stop_reason}")
print(f"Input tokens: {message.usage.input_tokens}")
print(f"Output tokens: {message.usage.output_tokens}")
print(f"Text: {message.content[0].text}")

Cell 3 — System Prompt: Give Claude a Persona

message = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=150,
    system="You are a cinematic director who speaks in short, vivid sentences.",
    messages=[{
        "role": "user",
        "content": "Describe a sunrise over a golf course."
    }]
)
print(message.content[0].text)

Cell 4 — Variables in Prompts

platform = "Kling"
shot = "slow motion golf swing with morning light"

message = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=200,
    messages=[{
        "role": "user",
        "content": f"Write a {platform} AI video prompt for: {shot}. "
                   f"Include camera, lighting, mood. Under 50 words. Only the prompt."
    }]
)

result = message.content[0].text
print(f"[{platform}] {result}")

Cell 5 — Error Handling

try:
    message = client.messages.create(
        model="claude-sonnet-4-6",
        max_tokens=100,
        messages=[{"role": "user", "content": "Hello"}]
    )
    print(f"Success: {message.content[0].text}")
except anthropic.AuthenticationError:
    print("Bad API key — check your .env file.")
except anthropic.RateLimitError:
    print("Too many requests — wait a minute and try again.")
except anthropic.APIError as e:
    print(f"API error: {e}")

End of Day Checklist

• pip install anthropic python-dotenv completed without errors
• Your API key is in a .env file — not in your Python code
• Cell 1 ran successfully and Claude responded
• You inspected the token usage in Cell 2
• You experimented with the system prompt in Cell 3
• You understand how f-strings power dynamic prompts in Cell 4
• You have error handling wrappers from Cell 5 ready to copy-paste