ESP-AI
The simplest and most cost-effective AI integration solution, enabling any object to have intelligent conversation capabilities. Let everyone have an AI assistant
Simplicity First
Never let a developer learn a new concept or write code that has nothing to do with the business. The client supports the use of Ardunio, IDF(in implementation), PlatformIO and other development environments.
Flexible Expansion
Built-in IFLYtek IAT, TTS, LLM, other platforms can be extended by plug-ins.
Voice Activation
Customizable offline voice activation with voice inference using TensorFlow Lite models. An online wake word generation service will be available in the future.
Command Recognition
Developers can customize user commands or use built-in commands to enable the robot to control home appliances, sing, and more.
C/S Architecture
The server and client are separated, and the complex calculation and service scheduling are all handed over to the server, which greatly reduces the cost of the client, and can provide services for N clients at the same time.
Identity Authentication
Each client connection and session can undergo identity verification and other operations.
Real-time Response
All service interfaces use the WebSocket protocol for streaming transmission, enabling real-time server response. Each character inferred by LLM is responded to in real time.
Configurable & Zero Code
Apart from plugins written for special business needs, the configurations provided by the framework or plugins generally meet most scenarios. A configuration generation service will be offered in the future, making it completely zero-code.
Production grade & integrable
ESP-AI is a development library released for production environments, with strict requirements for scalability, flexibility, and stability.