Realtime API demo for voice-driven 3D scene navigation
Top 68.1% on sourcepulse
This project demonstrates voice-based interaction with a 3D solar system model using OpenAI's Realtime API and WebRTC. It targets developers and users interested in integrating natural language control with interactive 3D environments, enabling dynamic scene manipulation and data visualization through conversational commands.
How It Works
The demo leverages OpenAI's Function Calling mechanism to interpret user voice commands. These commands are mapped to specific tools that trigger actions within a Spline-rendered 3D solar system. The Realtime API facilitates low-latency communication, enabling immediate visual feedback and scene updates based on spoken queries about planets, their data, or celestial events like the ISS's position.
Quick Start & Requirements
npm install
npm run dev
OPENAI_API_KEY
environment variable or a .env
file.Highlighted Details
Maintenance & Community
This is an official OpenAI demo project. Further community engagement or maintenance details are not specified in the README.
Licensing & Compatibility
Limitations & Caveats
The initial scene load can be slow due to its size. Voice recognition may be sensitive to background noise or echo. Customization requires familiarity with Spline's event system and modifying TypeScript configuration files.
6 months ago
1 day