Ai-Thinker Offline Voice Module VC-01 & VC-02 Series Tutorial — Fundamentals

[Yaqing-Zheng]

Introduction

An offline voice module is a speech recognition and voice control device that does not depend on the internet or any cloud service. It runs speech recognition, speech synthesis, and voice control locally, which makes it a good fit for IoT, smart home, robotics, embedded systems, and many other scenarios. Offline voice modules come with a wide range of characteristics, but the most notable ones are pre-defined voice commands and fast response times.

VC01 and VC02 Series Overview

The VC01 and VC02 series (including both the modules and their development board kits) are offline speech recognition modules from Ai-Thinker. Both are built around Unisound's Fengniao M (US516P6) speech recognition chip, offering high reliability and broad versatility.

These modules support analog microphone input and dual-channel DAC output, integrate 2MB Flash and 242KB SRAM, and feature multiple voice noise-reduction technologies that deliver a high recognition rate with low-latency response. They support both Chinese and English voice control, making them suitable for smart home, voice interaction, IoT, and similar applications.

The VC01 and VC02 dev kits share the same PCB base board — only the module package size differs (clearly visible in the image above). Both modules and their dev boards support up to 150 offline voice commands. Both also provide 5 GPIO outputs, I2C, PWM, DAC, and UART. The GPIO outputs can drive simple switches, while the other output protocols let you attach an external MCU for more complex command control.

By comparison, the VC02 is the better choice for users who are sensitive to board size. The dev board schematics can be found in the VC01 or VC02 dev board datasheets (see the image below).

Meanwhile, the offline voice module section of the Ai-Thinker docs site provides comprehensive support for the VC01 and VC02 series.

Unboxing the VC Dev Kit

Since the VC01-dev-kit uses the same base board as the VC02, the walkthrough below uses the VC02-dev-kit unless otherwise noted. The VC02-dev-kit contains three parts: the speaker module, the microphone, and the VC02-dev-kit board itself (see the image below).

The dev kit is powered through a Micro-USB port, and the board has two connector slots. The lower slot is the speaker output, and the upper slot is the microphone input. The image below shows the assembled board.

When connecting the microphone and speaker, make sure the connectors are fully seated. Thanks to the keyed (fool-proof) design, the connectors are easy to attach correctly. The image below shows the back of the dev kit.

Summary

In this article we briefly introduced offline voice modules and their use cases, and took a closer look at the Ai-Thinker VC01 and VC02 series modules and dev boards — covering their specification comparison and development resources. We also unboxed the VC02-Dev-Kit and highlighted its hardware features as a reference for developers. In the next article, I will cover the VC series pinout and peripheral expansion, along with testing the factory firmware.

---

[yakubumustapha270-a11y]

Please can you discuss on the creation of profile and reprogramming the vc-02.
Thanks 🙏🏾

[Yaqing-Zheng]

Hi yakubumustapha270, thanks for the question!

Yes, the next chapter in the series will cover exactly these two topics — how to create a custom profile (wake word + voice commands) on the Ai-Thinker Voice Open Platform, and how to re-flash the VC-02 with your custom firmware. I'll publish it on this Discussion board once it's ready, and tag you here when it's up.

Stay tuned!