AI-Thinker VC-02 Offline Speech Recognition Module

Description

The AI-Thinker VC-02 module integrates a dedicated offline speech recognition solution centered on the Unisound US516P6 voice processing chip. This chip employs a 32-bit RISC architecture enhanced with a digital signal processor (DSP) instruction set, a floating-point unit (FPU), and an FFT accelerator, enabling efficient, real-time audio analysis and command identification entirely on-device.

Designed for cost-sensitive and privacy-focused applications, the module runs a lightweight real-time operating system (RTOS) and supports a vocabulary of 150 pre-trained commands. Its peripheral suite includes UART (default 115200 bps), I2C, SPI, PWM, and multiple GPIOs, allowing direct control of relays, LEDs, or sensors. The module’s wide 3.6V to 5V supply range and industrial-grade operating temperature from -40°C to 85°C ensure reliability in diverse embedded environments.