XiaoZhi AI Voice Robot
Open-Source AI Voice Robot, Intelligence at Your Command!
Native XiaoZhi Dev Board Support | Zero-Code LLM+ASR+TTS Integration | Multilingual Dialogue + IoT Control
ESP32-based XiaoZhi AI Dev Board with Complete MCP Protocol Development Solutions
Core Features
· Offline Voice Wake-up · Multilingual ASR (CN/EN/JP/KR) · Real-time Voice Dialogue · LLM Integration (Qwen/DeepSeek/Doubao) · MCP Protocol IoT Control · Display & LED Feedback
Multi-Platform Compatibility
XiaoZhi firmware supports ESP32, ESP32-C3/C5/C6, ESP32-S3 and ESP32-P4; release v2.5.2 ships 188 prebuilt firmware variants across 154 board directories, ready for rapid secondary development
Technical Architecture
· HAL: Singleton-pattern Unified Interfaces · Audio Pipeline: Capture→Resample→Encode · Protocols: WebSocket/MQTT+UDP/MCP · AI Stack: Wake-up, Recognition, LLM Integration
Development & Deployment
· Select a supported board · Configure with ESP-IDF v6.0.1+ environment · Build & flash via idf.py command · Deploy supporting server-side programs
Design Highlights
· High-compatibility Development Board Design · Abstract Peripheral Support Classes · Dynamic Power Management
Hardware Forms
Breadboard DIY · Boards with Displays · AI Pendant and Other Forms · Wi-Fi / Ethernet / Cat.1 4G
Open-Source Ecosystem
MIT-licensed for commercial use with full customization support
Extensibility
· Custom Wake Words, Fonts, Emotions and Chat Backgrounds · On-device and Cloud MCP · Camera Vision on Selected Hardware · Speaker Recognition