Gemini Cursor ✨

A second AI cursor 🖱️ for your desktop that can see your screen, hear you speak, and talk to you.

Powered by Google's Gemini 2.0 Flash (Experimental) model, the Multimodal Live API, Pointing, and Function calling capabilities.

Created by @13point5.

Features

• 🖱️ Second AI cursor on your desktop
• 🚀 Multimodality: The model can see 📸, hear 🎤, and speak 🔊
• ⚡️ Real-time with low latency

Use Cases

• 📚 Understanding complex diagrams in Research papers, Architecture diagrams, etc
• 🌐 Navigating complex websites to perform a task like adding a payment method on Amazon
• 📝 Real time AI tutor with whiteboards

Tech Stack

• Frontend: Electron, React, TypeScript, Vite
• AI: Google Gemini API

Acknowledgements

• A lot of code from the Gemini Multimodal Live API Web console
• Built using Google's Multimodal Live API

Prerequisites

• Node.js (v16 or higher)
• npm
• Gemini API key

Installation

1. Clone the repository

git clone https://github.com/13point5/gemini-cursor.git
cd gemini-cursor

2. Install dependencies

npm install

3. Run the app

npm run start

4. Enter the Gemini API key in the app

5. Click the Play button and the Share Screen button

6. Minimize the app and enjoy!