Ask me anything
Ithy Logo
Sign in to save

Getting Started with Google Generative AI (GenAI) Library in Python

[Sponsored] Google Cloud Post: IBL Education’s GenAI-Based Chat Mentor ...

This guide provides a comprehensive walkthrough on how to set up and use the Google Generative AI (GenAI) library in Python. It covers everything from initial setup to advanced features, enabling you to integrate Google's powerful generative models into your applications.

Prerequisites

Before you begin, ensure you have the following:

  • Python Version: Python 3.7 or higher is required. You can check your version by running python --version in your terminal. If needed, download the latest version from python.org.
  • Google Cloud Account: You need a Google Cloud account to access the Generative AI API. If you don't have one, sign up at cloud.google.com.
  • API Key: You'll need an API key to authenticate your requests to the Google GenAI API. This key can be generated in the Google Cloud Console.
  • Basic Python Knowledge: A basic understanding of Python programming is necessary to follow along with the setup and usage of the library.

Setting Up Your Environment

It's recommended to use a virtual environment to manage dependencies for your project. This helps avoid conflicts with other Python projects. Here's how to set one up:

  1. Create a Virtual Environment:

    Use the following command to create a virtual environment named genai_env:

    python -m venv genai_env
  2. Activate the Virtual Environment:

    On macOS and Linux, use:

    source genai_env/bin/activate

    On Windows, use:

    genai_env\Scripts\activate

Installing Required Packages

The primary package you need is google-generativeai, which allows you to interact with Google's Generative AI models. You should also install python-dotenv for managing your API key securely, and optionally Pillow if you plan to work with images.

Install these packages using pip:

pip install -U google-generativeai python-dotenv Pillow

Configuring Your API Key

To use the Google Generative AI API, you need to configure your API key. Here’s how you can do it:

  1. Generate an API Key:
    • Go to the Google Cloud Console.
    • Create a new project or select an existing one.
    • Navigate to APIs & Services > Library and enable the "Google Generative AI API."
    • Go to APIs & Services > Credentials and click Create Credentials > API Key.
    • Copy the generated API key.
  2. Create a .env File:

    In your project directory, create a file named .env.

  3. Add Your API Key:

    Open the .env file and add your Google API key as follows:

    GOOGLE_API_KEY=your_google_api_key_here

    Replace your_google_api_key_here with your actual API key. This ensures your API key remains secure and is not hard-coded in your scripts.

Setting Up the Generative Model

Next, you need to import the required libraries and configure the Generative AI model using your API key.


import google.generativeai as genai
import os
from dotenv import load_dotenv

# Load API key from .env file
load_dotenv()
google_api_key = os.getenv('GOOGLE_API_KEY')

# Configure the Generative AI model
genai.configure(api_key=google_api_key)

This code loads the API key from the .env file and configures the google-generativeai library to use this key.

Generating Content

With the model set up, you can now generate content using the API. Here’s a simple example to get you started:


# Create an instance of the GenerativeModel
model = genai.GenerativeModel('gemini-pro')

# Generate content based on a prompt
response = model.generate_content("Hello, can you introduce yourself?")
print(response.text)

This code generates a friendly introduction based on the provided prompt and prints the response.

Implementing Automated Conversations

To create an interactive automated conversation, you can use a loop that continuously sends user inputs to the model and prints the responses. Here’s how you can do it:


# Create an instance of the GenerativeModel
model = genai.GenerativeModel('gemini-pro')

# Start a chat session
chat = model.start_chat()

def automated_conversation():
    while True:
        user_input = input("You: ")
        if user_input.lower() == "exit":
            print("Conversation ended.")
            break
        response = chat.send_message(user_input)
        print("AI:", response.text)

# Start the automated conversation
automated_conversation()

This script initiates a chat session and allows for an ongoing conversation with the AI until the user types “exit”.

Working with Images

If you want to use the GenAI library for image-related tasks, you can load and process images using the Pillow library. Here’s an example:


from PIL import Image

# Load an image
image_path = "path/to/your/image.jpg"
image = Image.open(image_path)

# Define a prompt for the image
prompt = "Describe the content of this image."

# Send the image and prompt to the API
response = model.generate_content(
    contents=[prompt, image],
)

# Print the response
print("Image Analysis:")
print(response.text)

Full Example

Here is a full example that includes all the steps mentioned above:


import google.generativeai as genai
import os
from dotenv import load_dotenv
from PIL import Image

# Load API key from .env file
load_dotenv()
google_api_key = os.getenv('GOOGLE_API_KEY')

# Configure the Generative AI model
genai.configure(api_key=google_api_key)

# Create an instance of the GenerativeModel
model = genai.GenerativeModel('gemini-pro')

# Generate content based on a prompt
response = model.generate_content("Hello, can you introduce yourself?")
print(response.text)

# Start a chat session
chat = model.start_chat()

def automated_conversation():
    while True:
        user_input = input("You: ")
        if user_input.lower() == "exit":
            print("Conversation ended.")
            break
        response = chat.send_message(user_input)
        print("AI:", response.text)

# Start the automated conversation
automated_conversation()

# Load an image
image_path = "path/to/your/image.jpg"
try:
    image = Image.open(image_path)
    # Define a prompt for the image
    prompt = "Describe the content of this image."

    # Send the image and prompt to the API
    response = model.generate_content(
        contents=[prompt, image],
    )

    # Print the response
    print("Image Analysis:")
    print(response.text)
except FileNotFoundError:
    print(f"Error: Image file not found at {image_path}")
except Exception as e:
    print(f"An error occurred: {e}")

This example sets up the environment, configures the API key, generates content based on a prompt, initiates an automated conversation, and demonstrates image analysis.

Advanced Features

Streaming Responses

For long responses, you can stream the output:


response = model.generate_content(
    "Explain the theory of relativity in simple terms.",
    stream=True
)

for chunk in response:
    print(chunk.text, end="")

Token Counting

To estimate the number of tokens in a prompt:


token_count = genai.tokenize("This is a sample prompt.")
print(f"Token Count: {len(token_count)}")

Safety Settings

Enable safety settings to filter inappropriate content:


response = model.generate_content(
    "Generate a story for children.",
    safety_settings={"category": "violence", "threshold": 0.5}
)
print(response.text)

System Instructions

You can provide system instructions to guide the model's behavior. For example, you can instruct the model to respond in a specific tone or style:


chat = model.start_chat()
response = chat.send_message("Explain quantum computing in simple terms.", system_instruction="You are a helpful and friendly AI assistant.")
print(response.text)

Function Calling

The library supports function calling, allowing you to pass Python functions directly to the model. The model can then call these functions and respond accordingly:


def add_numbers(a, b):
    return a + b

response = chat.send_message("What is 2 + 2?", functions=[add_numbers])
print(response.text)

Asynchronous API Calls

For applications requiring asynchronous operations, use Python’s asyncio library:


import asyncio

async def main():
    # Define the model and input prompt
    model_id = "gemini-pro"
    prompt = "What are the latest trends in AI?"

    async with genai.client.aio.live.connect(model=model_id) as session:
        await session.send(prompt, end_of_turn=True)

        # Process responses
        async for response in session.receive():
            if not response.server_content.turn_complete:
                for part in response.server_content.model_turn.parts:
                    print(part.text, end="", flush=True)

asyncio.run(main())

Error Handling

Always handle errors gracefully:


try:
    response = model.generate_content(
        "Generate a motivational quote.",
    )
    print(response.text)
except Exception as e:
    print(f"An error occurred: {e}")

Additional Tips and Resources

Conclusion

By following this guide, you can effectively set up and use the Google Generative AI library in your Python applications. This enables you to generate content, engage in automated conversations, analyze images, and more. Experiment with different models, prompts, and configurations to unlock the full potential of Google Generative AI. Happy coding!


Last updated January 6, 2025

AI-assisted research. Sources, methodology and corrections · Source-checked guides and tools · Explore published answers

Ask Ithy AI