
BeagleY-AI
Raspberry Pi HAT-compatible edge AI SBC by BeagleBoard.org, powered by the TI AM67A SoC with 4 TOPS neural accelerator, 4GB LPDDR4, 16GB eMMC, WiFi, and Gigabit Ethernet.
$70.00
Tap any value to browse matching devices
Specifications
Components
TI AM67A quad-core Cortex-A53 @ 1.4GHz with C7x DSP and 4 TOPS MMA neural accelerator for edge vision applications.
Overview
The BeagleY-AI is a Raspberry Pi HAT-compatible single-board computer from the BeagleBoard.org Foundation, designed to bring serious edge AI capabilities to the familiar 40-pin form factor. It is powered by the Texas Instruments AM67A SoC, which combines a quad-core ARM Cortex-A53 cluster with a TI C7x DSP vector processor and a 4 TOPS MMA neural network accelerator for hardware-accelerated machine learning inference.
Equipped with 4GB of LPDDR4 RAM and 16GB of eMMC storage alongside a microSD slot, the BeagleY-AI provides ample memory and storage for demanding computer vision and AI workloads. Wireless connectivity includes WiFi 5 and Bluetooth 5.0 alongside Gigabit Ethernet, making it easy to integrate into network-connected AI applications. The USB-C 3.0 port supports fast peripheral connectivity and board power.
The HAT compatibility means developers can immediately leverage the vast Raspberry Pi accessory ecosystem for cameras, displays, sensors, and industrial interfaces. Running Armbian or the official TI-based Debian image, the BeagleY-AI bridges the gap between hobby-grade SBCs and production-grade AI development platforms, making it an excellent choice for vision AI projects, robotics, and edge inference applications.
Use Cases
4Edge AI
On-device machine learning inference at the edge, without cloud dependency
Computer Vision
Image recognition, object detection, and ML inference using embedded cameras and AI-capable processors.
Robotics
General robotics applications including mobile robots, robotic arms, drones, and automated systems.
rpi-hat-projects
Where to Buy
Explore more of the directory — popular use cases, firmware, and specs to browse next.




