Easy AI Module

A simple Python package for personal use that makes it easy to interact with Google's Gemini AI API.

Installation

pip install -e .

Usage

First, replace the API_KEY in easy_ai_module/gemini_client.py with your actual Gemini API key.

Text Generation

from easy_ai_module import GeminiAI

ai = GeminiAI()  # Defaults to gemini-2.0-flash

# Generate text from a prompt
response = ai.generate_text("Write a short story about a robot")
print(response)

# Customize parameters
response = ai.generate_text(
    prompt="Write a haiku about the moon",
    max_tokens=100,
    temperature=0.3  # Lower temperature for more focused output
)
print(response)

Image Analysis

from easy_ai_module import GeminiAI

ai = GeminiAI()

# Analyze an image (requires valid image URL)
image_url = "https://example.com/image.jpg"
response = ai.generate_with_image("What's in this image?", image_url)
print(response)

Models

You can specify different Gemini models:

# Use Gemini 1.5 Flash
ai = GeminiAI(model="gemini-1.5-flash")

# Use Gemma 3
ai = GeminiAI(model="gemma-3")

# Use Gemini 2.0 Flash (default)
ai = GeminiAI(model="gemini-2.0-flash")

# Use Gemini 2.0 Flash-Lite
ai = GeminiAI(model="gemini-2.0-flash-lite")

Model Selection Guide

The package includes a helpful guide to select the right model for your needs:

# Show general help on model selection
from easy_ai_module import GeminiAI
GeminiAI.help()

# Get recommendation for a specific use case
GeminiAI.help('reasoning')  # For complex reasoning tasks
GeminiAI.help('speed')      # For fastest response times
GeminiAI.help('budget')     # For cost-effective usage
GeminiAI.help('creative')   # For creative content generation
GeminiAI.help('image')      # For image analysis tasks
GeminiAI.help('local')      # For models that can run locally

Detailed Model Information

For detailed information about model capabilities, strengths, and pricing:

from easy_ai_module import print_model_guide

# Show information about all models
print_model_guide()

# Show information about a specific model
print_model_guide('gemini-2.0-flash')

Testing Tools

The package includes tools for testing and benchmarking models:

# Run a simple test with the default model
python model_demo.py

# Show the detailed guide for all models
python model_demo.py --guide

# Show help for a specific use case
python model_demo.py --show-help --use-case creative

# Run benchmarks on all models
python model_demo.py --benchmark

# Test a specific model with a custom prompt
python model_demo.py --model gemini-2.0-flash-lite --prompt "Explain quantum computing"

# Run all tests (help menu, benchmarks, and model guide)
python model_demo.py --test-all --verbose

Note

This package is for personal use only and contains a hardcoded API key. Do not share it with others.