index.md

layout	home
hero	name text tagline actions MetaHuman Engine Browser-Native 3D Digital Human Engine Complete interaction loop: Voice Input → AI Dialogue → Voice Output → 3D Expression Animation theme text link brand Get Started /en/guide/getting-started theme text link alt API Reference /en/api/overview
features	icon title details 🎭 3D Avatar High-quality 3D rendering with Three.js + React Three Fiber, supporting facial expressions and gestures icon title details 🗣️ Voice Interaction Web Speech API for recognition + Edge TTS for synthesis, multi-language voice support icon title details 👁️ Vision Sensing MediaPipe face mesh and pose detection, real-time expression capture to drive the avatar icon title details 🧠 AI Dialogue OpenAI-compatible API interface, streaming responses and context management

MetaHuman Engine Documentation

Core Capabilities

Capability	Technology	Status
🎭 3D Avatar	Three.js + React Three Fiber	✅ Available
🗣️ Voice	Web Speech API + Edge TTS	✅ Available
👁️ Vision	MediaPipe Face Mesh & Pose	✅ Available
🧠 Dialogue	OpenAI-compatible API	✅ Available

Quick Links

• Getting Started - Up and running in 5 minutes
• Installation Guide - Detailed setup instructions
• Configuration - Environment variables reference
• API Documentation - Backend API reference
• Architecture - System design

Design Principles

1. Zero-config by default — Works out of the box with automatic fallback to local mode
2. Graceful degradation — External service failures don't break the core experience
3. Modular architecture — Avatar, voice, vision, dialogue are independent and replaceable
4. Minimal re-renders — Focused state stores prevent unnecessary updates
5. Browser-first — Process client-side whenever possible, minimize server dependency
6. Privacy-first — Face data never leaves the browser