Offline AI Chatbot

A desktop application that provides a clean interface for interacting with Ollama's AI models locally. Chat with AI models without needing internet connectivity after initial setup.

Features

• Fully offline AI chat capabilities
• Multiple AI model support through Ollama
• Support for basic text files, CSVs, and JSON
• Support for images compatible with select models
  • LLaVa and Llama 3.2 for example
• Clean, modern interface
• Dark/Light mode support
• Real-time responses
• Local data storage

Prerequisites

1. Install Ollama:
   • Visit Ollama's website
   • Download and install for your system
   • Open terminal and verify installation:

     ollama --version
   • Pull and run your first model:

     ollama pull llama2
     ollama run llama2
   • Check out my blog post for more information on how to get started with Ollama.

Setup

1. Clone the repository

   git clone https://github.com/yourusername/offline-chatbot.git
   cd offline-chatbot

2. Install dependencies

   npm install

3. Create a .env file in the root directory:

   VITE_PORT=3030

Running the Application

Start both frontend and backend servers:

npm start

This will run:

• Frontend: http://localhost:5173
• Backend: http://localhost:3030

Development

Run frontend only:

npm run dev

Run backend only:

npm run server

Project Structure

offline-chatbot/
├── src/               # Frontend source code
├── server/            # Backend server code
└── public/            # Static assets

Environment Variables

Variable	Description	Required
VITE_PORT	Backend server port	Yes

Tech Stack

• React + Vite
• Express.js
• Ollama API
• TailwindCSS
• Node.js

Additional Resources

• Offline Chatbots with Ollama - Blog Post
• Ollama JS Documentation
• Ollama Model Library

License

MIT