AI Instructor Live Labs Included

GCP: Gemini API Fundamentals

Build real applications with Google's Gemini API in Python — text generation, vision, chat, and function calling from scratch.

Beginner

9h 0m

6 Lessons

About This Course

Master the Google Gemini API using Python. Learn to generate text, work with multimodal inputs (images, PDFs), build multi-turn chat applications, and implement function calling. All exercises use the Python AI container with the google-generativeai SDK and real API calls against Gemini 2.0 Flash.

Course Curriculum

6 Lessons

AI Lesson

GCP: Introduction to Gemini & the Gemini API

45m

Learn what Gemini is, explore the model family (2.0 Flash, 1.5 Pro, 1.5 Flash), understand the API architecture, authentication with API keys, safety settings, and pricing model.

Lab Exercise

GCP: Your First Gemini API Calls

1h 55m 4 Exercises

Set up the Python SDK, configure your API key, generate text with Gemini 2.0 Flash, tune model parameters, and stream responses — all from the VS Code terminal with real API calls.

Set Up the SDK and Verify Your API Key Verify the google-generativeai SDK is installed, load your API key from the environment, and confirm you can connect to the Gemini API by listing available models. ~15 min

Generate Text with Gemini 2.0 Flash Send your first prompts to Gemini 2.0 Flash and work with the response object — accessing text, token counts, and finish reason. ~20 min

Tune Model Parameters Control Gemini's output using GenerationConfig — experiment with temperature, max_output_tokens, top_p, and top_k to see how each parameter shapes the response. ~20 min

Stream a Response Use streaming mode to receive Gemini's response token-by-token, printing output as it arrives — the foundation for building responsive AI-powered applications. ~15 min

AI Lesson

GCP: Multimodal & Advanced Gemini Features

45m

Learn how Gemini processes images and other media types, build multi-turn chat conversations with history, implement function calling to connect Gemini to your own Python functions, and use system instructions and JSON output mode.

Lab Exercise

GCP: Building a Gemini Chat Application

2h 25m 4 Exercises

Build a complete multi-modal chat application using the Gemini API. You'll implement multi-turn conversations, image analysis, function calling, and combine everything into a production-ready chat app with streaming.

Multi-Turn Conversation with History Implement a persistent chat session using Gemini's multi-turn conversation API. You'll build the chat session factory and message-sending functions that maintain conversation history across turns. ~25 min

Analyze Images with Gemini Vision Use Gemini's multimodal capabilities to analyze images. You'll implement functions that describe images, answer questions about them, and compare multiple images — all by passing PIL Image objects alongside text prompts. ~20 min

Implement Function Calling Give Gemini the ability to call Python functions. You'll define FunctionDeclaration schemas for weather lookup and currency conversion, then implement the two-turn cycle: detect function_call → execute Python → send FunctionResponse back to the model. ~25 min

Build the Complete Chat Application Combine everything into a fully-featured chat app with streaming output. The app supports slash commands (/image, /weather, /convert, /history, /clear) and routes each to the appropriate Gemini capability you built in the previous exercises. ~30 min

AI Lesson

GCP: AI Image Generation — Nano Banana 2 and Pro

45m

Learn how Google's Nano Banana image generation models work — Nano Banana 2 (gemini-3.1-flash-image-preview) for fast, high-volume generation and Nano Banana Pro (gemini-3-pro-image-preview) for maximum quality with built-in thinking. Understand the image generation API, response_modalities, ImageConfig parameters, and when to choose each model.

Lab Exercise

GCP: Hands-On Image Generation with Nano Banana

2h 25m 4 Exercises

Use the Gemini image generation API in VS Code to generate images from text prompts, edit reference images, build a multi-turn image editing session, and compare the quality output of Nano Banana 2 versus Nano Banana Pro. All exercises use the new google-genai SDK.

Generate Images from Text with Nano Banana 2 Use the Gemini image generation API with Nano Banana 2 (gemini-3.1-flash-image-preview) to generate images from text prompts. You'll install the dependencies, set up your API key, generate your first image, experiment with aspect ratios and sizes, and save images to disk. ~25 min

Edit Images with Reference Input Pass a reference image alongside an edit instruction to Nano Banana 2. You'll load a downloaded reference image, send it to the model with various editing prompts (background change, style transfer, object addition), and compare the edited outputs. ~25 min

Multi-Turn Image Editing with Chat Sessions Use client.chats.create() with an image generation model to build a multi-turn editing workflow. Generate an initial scene, then iteratively refine it across three conversation turns, with each turn building on the model's memory of the previous image. ~25 min

Nano Banana Pro — High-Fidelity Generation and Quality Comparison Switch to Nano Banana Pro (gemini-3-pro-image-preview) and generate the same prompts used in Exercise 1. Compare the outputs side-by-side to see the quality difference, use 4K resolution (Pro-only), and understand when the quality uplift justifies the cost. ~25 min

This course includes:

24/7 AI Instructor Support
Live Lab Environments
3 Hands-on Lessons
6 Months Access
Certificate of Completion