3.1.1: Welcome & First Generation

• Time to Complete: 10 minutes
• Prerequisites: Nano Banana files downloaded (Module 3.0)

Start this module in Cursor: Run /start-3-1-1 to begin the interactive experience.

Overview

Module 3.1.1 gets you from zero to your first AI-generated image. You’ll set up your Gemini API key, configure billing, and watch as the AI transforms a reference photo into something magical.

Key takeaway: The setup is a one-time process. Once your API key is configured, you can generate images with simple natural language requests - The AI handles all the technical complexity.

What is Gemini 3 Pro?

Gemini 3 Pro (codenamed “Nano Banana Pro”) is Google’s most advanced image generation model. It can:

• Generate photorealistic images from text descriptions
• Transform reference photos while preserving identity
• Add accurate text overlays to images
• Create consistent characters across multiple generations
• Understand and replicate complex visual styles

Unlike consumer apps, accessing the API directly gives you:

• More permissive content policies (e.g., real people as references)
• Full parameter control (aspect ratio, resolution, etc.)
• Session-based iteration (refine images step-by-step)
• Lower cost (~$0.10/image vs subscription apps)

Setting Up Your API Key

Step 1: Get Your Key from Google AI Studio

1. Go to Google AI Studio
2. Sign in with your Google account
3. Accept the Terms of Service (if first time)
4. Click “Get API Key” in the left sidebar
5. Click “Create API key” in the upper right to generate a new key
6. Name your key whatever you want and select “Default Gemini Project”
7. Click on the key in the dashbaord
8. Copy the key - it starts with AIza...

Keep this key safe. You’ll paste it into your .env file in Step 3.

Step 2: Set Up Billing

Gemini 3 Pro requires a billing account. Don’t worry - it’s pay-per-use and very affordable.

1. In Google AI Studio, go to Get API key (bottom of left sidebar)
2. Under the “Quota tier” column, click Set up billing
3. Follow the prompts to add a payment method

Cost context:

• ~$0.10 per image generation
• This entire course costs less than $5 total
• No monthly subscription - just pay for what you use
• This money goes to Google, not Carl (unfortunately)

Step 3: Configure Your API Key

Just tell the AI your API key and it will set everything up:

“My Gemini API key is AIzaYourKeyHere - please add it to my .env file”

The AI will create the .env file and store your key securely.

Prefer to do it manually? Copy .env.example to .env and add your key: GEMINI_API_KEY=YourKeyHere

Your First Generation

Once setup is complete, you’re ready to generate. In the interactive module, The AI uses a reference photo to create a fun welcome image - demonstrating the model’s ability to transform real photos.

What happens behind the scenes:

1. You describe what you want
2. The AI calls the generate() function from image_gen.py
3. The function sends your request to Gemini’s API
4. The generated image is saved to the outputs/ folder
5. The AI tells you where to find it

Generation time: About 10-15 seconds per image.

What the AI Handles for You

When you generate images through Cursor, you don’t need to worry about:

This is the power of combining Gemini with Cursor. You focus on creative direction; The AI handles implementation.

The image_gen.py Module

All image generation runs through a pre-built Python module called image_gen.py. You don’t need to write any code - The AI uses it on your behalf.

Key functions:

• generate() - Create or refine an image
• new_session() - Start fresh (clear conversation history)
• session_info() - Check current session status

You’ll learn more about these in Module 3.1.2.

Output Structure

Generated images are saved to the outputs/ folder in your nano-banana directory:

nano-banana/
├── outputs/
│   ├── image_001.png
│   ├── image_002.png
│   └── ...
├── image_gen.py
├── .env
└── ...

Images are numbered sequentially. The AI will tell you the exact path to each generated image so you can open it in Finder (Mac) or Explorer (Windows).

Troubleshooting

”API key not found” or authentication errors

• Verify your .env file exists (not .env.example)
• Check that the key is pasted correctly with no extra spaces
• Make sure the key starts with AIza

”Billing not enabled” errors

• Return to Google AI Studio → Settings → Plan information
• Confirm billing is set up and a payment method is linked
• It may take a few minutes for billing to activate

Generation fails or times out

• Check your internet connection
• Gemini may be experiencing high load - wait a minute and try again
• If errors persist, try starting a new session: tell the AI “Start a new session”

Can’t find generated images

• Images save to the outputs/ folder
• The AI will provide the exact path after each generation
• Open Finder (Mac) or Explorer (Windows) and navigate to the folder

Best Practices

Keep your API key secure:

• Never commit .env to git (it’s in .gitignore by default)
• Don’t share your API key publicly
• You can regenerate keys in Google AI Studio if compromised

Check outputs in a proper viewer:

• Terminal can’t display images - always open the file
• Use Preview (Mac), Photos (Windows), or any image viewer
• For best experience, keep the outputs/ folder open while working

What You’ll Learn Next

Module 3.1.1 gets you set up. The rest of Module 3.1 teaches you to use image generation effectively:

• 3.1.2: Understanding the Basics - Parameters, aspect ratios, resolution, iteration
• 3.1.3: Consistency & Style - Golden Rules of prompting, reference images, variants
• 3.1.4: Building Your Style Database - Create a reusable library of styles

Then Module 3.2 applies these skills to real PM work: personas, diagrams, mockups, and marketing assets.

What’s Next?

You’re set up and you’ve generated your first image. Now it’s time to understand how the system works.

Module 3.1.2 teaches you about the generate() function, available parameters, and how to iterate on your images.

Interactive track: Type /start-3-1-2

Resources

• Google AI Studio - Get your API key and manage billing
• Gemini Image Generation Documentation - Official API reference

About This Course

Created by Carl Vellotti. Check out The Full Stack PM for more PM builder content.

Source Repository: github.com/carlvellotti/claude-code-pm-course