Hiwonder WonderMK K230 Vision Module 6TOPS Equivalent AI Performance, Large Model Development Board with CanMV Camera

The Hiwonder WonderMK K230 Vision Module is a high-performance AI development board powered by a dual-core RISC-V processor with 6 TOPS equivalent computing power. It is an ideal platform for South African students, educators, and makers looking to explore advanced computer vision, multimodal large language models, and edge AI applications. With its built-in touchscreen, HD camera, and extensive controller compatibility, this module simplifies the deployment of complex robotics and machine learning projects.

Dual-core RISC-V processing power

Equipped with a dual-core RISC-V K230 processor and a 6 TOPS equivalent KPU, this module delivers approximately 13.7 times the performance of the older K210 chip. This massive boost allows for rapid edge computing and smooth execution of demanding AI workloads.

Multimodal large language model integration

The module supports cloud-based LLM APIs such as Alibaba Qwen, allowing seamless switching between text, speech, and vision models. This enables robots to achieve enhanced perception, reasoning, and real-time visual understanding for advanced embodied intelligence.

Simplified one-click AI training

A built-in user interface supports one-click AI training, colour sampling, and multi-face recognition. This removes complex machine learning workflows, significantly shortening your project development cycles.

Extensive computer vision library support

Fully compatible with MicroPython and official OpenMV libraries, the module supports over 30 AI vision functions including YOLOv8n and YOLOv11n object detection. It also allows for custom model training and includes comprehensive sample code to accelerate development.

High-definition digital camera

The onboard 1080p digital camera operates at 60 frames per second, providing significantly clearer imaging than standard OV5640 or OV5647 sensors. This high-resolution capture ensures superior accuracy for visual recognition tasks.

Interactive capacitive touchscreen

A 2.4-inch capacitive touchscreen with a 240x320 resolution provides an intuitive user interface directly on the module. This allows for real-time monitoring, debugging, and direct user interaction without needing an external display.

Versatile hardware expansion

Onboard I2C, UART, and 10-pin expansion interfaces allow the module to connect seamlessly with Arduino, Raspberry Pi, ESP32, and STM32 controllers. It also supports external speakers to enable voice interaction alongside real-time visual feedback.

Generous high-speed memory

Packed with 1GB of LPDDR4 memory, this board is well-equipped to handle high-memory scenarios. This ensures smooth performance when running complex AI vision applications and large models.