Introduction to the Google GenAI SDK

A concise cookbook demonstrating core capabilities of the official google-genai Python SDK using Gemini 2.5 models in Google Colab.

Features Covered

• Core Generation: Basic text generation, system instructions, and configuration tuning (temperature/top_p).
• Multimodal Inputs: Analyzing raw PIL images, online image URLs, and large multi-page PDF files.
• Advanced Workflows: Multi-turn chat sessions, streaming responses, and asynchronous API execution.
• Structured Outputs: Parsing text straight into JSON format using Pydantic models and explicit JSON schemas.
• Safety & Utilities: Implementing safety filters, token counting, and tool calling configurations.

Prerequisites

• A Google Colab notebook environment.
• A free developer API key from Google AI Studio.

Quick Setup

1. In Colab, click the Secrets (Key Icon) tab in the left sidebar.
2. Add a new secret named genai-sdk-gemini-api-key and paste your API key.
3. Toggle Notebook access to ON.
4. Run the code cells sequentially.

License

This project is open-source and licensed under the MIT License.