Mastering Google GenAI with Python

Google GenAI provides powerful tools for developers to integrate generative AI into their applications using Python. Whether you are building a chatbot, generating content, or creating embeddings, the Google GenAI library makes it easy.

In this article, we will explore how to set up and use Google GenAI with Python. We will cover installation, basic usage, and practical examples.

Prerequisites

Before diving into Google GenAI, ensure you have the following:

• A Google account
• Python 3.8 or higher installed
• Basic knowledge of Python programming
• An API key from the Google Cloud Console

If you are new to APIs, consider reading our Python Google Drive API Guide for Beginners to understand how APIs work.

Installing the Google GenAI Library

To start using Google GenAI, install the official Python package:

pip install google-genai

This command installs the latest version of the library. Keep it updated:

pip install --upgrade google-genai

Setting Up Your API Key

Generate an API key from the Google Cloud Console. Navigate to the APIs & Services section and create a new project.

Enable the Generative Language API for your project. Then create credentials and copy your API key.

Store your API key securely. You can set it as an environment variable:

export GOOGLE_API_KEY=your_api_key_here

Initializing the Client

Use the genai.Client class to initialize the client:

from google import genai

client = genai.Client(api_key="your_api_key_here")

The client handles authentication and communication with Google's servers.

Generating Text with a Model

Choose a model like gemini-1.5-flash for fast text generation:

response = client.models.generate_content(
    model="gemini-1.5-flash",
    contents="Explain quantum computing in simple terms."
)

print(response.text)

Example output:

Quantum computing uses qubits... [truncated for brevity]

The generate_content method returns a response object. Access the generated text via response.text.

Working with Embeddings

Convert text into numerical vectors using embeddings:

result = client.models.embed_content(
    model="text-embedding-004",
    contents=["Hello, world!", "How are you?"]
)

print(result.embeddings[0].values[:5])  # Print first 5 values

Output example:

[0.0123, -0.0456, 0.0789, -0.0234, 0.0678]

The embed_content method returns vector representations useful for similarity searches.

Listing Available Models

Check what models are available:

for model in client.models.list():
    print(model.name)

Output:

models/gemini-1.5-flash
models/gemini-1.5-pro
models/text-embedding-004
...

Use client.models.list to discover supported models.

Error Handling Best Practices

Always wrap API calls in try-except blocks:

try:
    response = client.models.generate_content(
        model="gemini-1.5-flash",
        contents="Hello!"
    )
    print(response.text)
except Exception as e:
    print(f"Error: {e}")

Conclusion

Google GenAI simplifies integrating advanced AI capabilities into Python applications. With proper setup and secure API management, developers can leverage state-of-the-art models for text generation and embeddings.

Start experimenting with small prompts and gradually build more complex workflows. Always follow security best practices when handling API keys and user data.