Chatter: A REPL for connecting to Google Gemini

A simple, text-based REPL (Read-Evaluate-Print Loop) application that connects to the Gemini API using Java and Python.

Project Goals

Demonstrate hybrid development using the strengths of Java and Python.
Explore basic API interaction with the Gemini AI platform.
Provide a foundation for building more sophisticated AI-powered text-based applications.

Prerequisites

Java Development Kit (JDK) – Installation instructions vary based on your OS.
Python 3 – Download from https://www.python.org/downloads/
google-generativeai Python library: Install using pip install google-generativeai
A Gemini API key: Obtain from https://ai.google.dev (sign-up may be required)

Setup

Clone or download this repository.
Set up your Gemini API key: Create a file named api_key.txt in the project's root directory and place your API key inside. Ensure this file is not committed to version control.
Install the google-generativeai library in your Python environment:
pip install google-generativeai

Running the Application

Navigate to the directory in your terminal.
Compile the code using javac Chatter.java
Run the program using java Chatter

Usage

Type in a prompt or a question and press Enter. The REPL will send your input to the Gemini API and display the response. Type 'goodbye' to exit the REPL.

Notes

This project is intended for demonstration and learning purposes.
For production applications, explore more robust API key management and error handling.
Consider using tools like gRPC for potentially more efficient communication between Java and Python processes.

Potential Enhancements

Improve the user interface with a Java library like Swing or JavaFX.
Add advanced error handling and input validation.
Implement a command-line argument parser for customization options.

Disclaimer

This project is not directly affiliated with or endorsed by Gemini AI or Google. Use the Gemini API in accordance with their terms of service.

sjweierman / chatter