March 17, 2025, San Francisco, CA – OpenMV is excited to announce the launch of the OpenMV N6 and OpenMV AE3, two groundbreaking machine vision cameras designed to bring real-time AI capabilities to microcontrollers. Backed by years of expertise in embedded vision, OpenMV is making high-performance AI vision accessible to developers, researchers, and hobbyists alike through its latest Kickstarter Campaign.
Revolutionizing Embedded AI with OpenMV N6
The OpenMV N6 is OpenMV’s most powerful camera yet, featuring a built-in Neural Processing Unit (NPU) capable of running real-time AI models like YOLO object detection at 30+ FPS at less than 0.75W. Powered by the STM32N6, the OpenMV N6 brings the power of a single-board computer to the microcontroller world with an 800 MHz Cortex-M55 (CM55) Vector Accelerated CPU, 1 GHz NPU (600 Gigaops), 1MP 120 FPS Color Global Shutter Camera, Image Signal Processor (ISP), Hardware H.264/JPEG Encoding, WiFi/BLE, Gigabit Ethernet, a GPU for image scaling, and extensive GPIO support. Whether tracking objects, monitoring security, or integrating AI into industrial automation, the OpenMV N6 is designed for professionals seeking high performance in a low-power package that is highly extensible.
Tiny, Mighty, and Ultra-Efficient: The OpenMV AE3
For those looking for a smaller, even lower-power AI vision solution, the OpenMV AE3 delivers in an ultra-compact form factor (1”x1”). Built around the Alif Ensemble E3 microcontroller, it features 400/160 MHz dual-core CM55 processing, dual Ethos-U55 NPUs (250 Gigaops), and a high-frame-rate 1MP Color Global Shutter Camera. With an ultra-low power draw of <60mA @ 5V (0.25W) at full power and <2.5mW in deep sleep, the AE3 is designed to run AI models like YOLO at ~30 FPS when awake while operating for years on a battery sleeping. It also features an IMU, microphone, an 8×8 distance sensor, and WiFi/BLE onboard.
Machine Vision and AI for Everyone – Powered by Python
OpenMV has always prioritized accessibility, and the N6 and AE3 continue this tradition by running MicroPython, allowing users to script powerful machine vision applications with just a few lines of Python code. Users program their OpenMV Cam through OpenMV IDE which makes it easy to load AI models, debug scripts, and visualize real-time video feeds. Additionally, the ULAB Module onboard provides fast NumPy-like data processing capabilities for handling AI model pre- and post-processing. The high-level Python interface, combined with onboard AI accelerators and vector processing extensions on the CPU, delivers over 200 times more ML performance on the OpenMV AE3 than previous OpenMV Cam models and up to 600 times more for the OpenMV N6.
Industry-Leading AI Training Integration
To simplify AI model training, OpenMV has partnered with Roboflow and Edge Impulse, providing seamless workflows for training and deploying AI models directly to the N6 and AE3. Kickstarter backers will receive three free months of Roboflow’s Basic Plan for training AI models like YOLO for Object Detection, Pose Estimation, and Segmentation. Edge Impulse also offers a platform for training AI models for the accelerometer, gyroscope, and microphone sensors of the OpenMV N6 and AE3, as well as for developing high-speed object tracking models like FOMO, which can perform object detection at over 200 FPS.
Expandable, Customizable, and Built for Real-World Applications
The OpenMV ecosystem includes powerful accessories and shields, such as thermal camera modules (FLIR® Lepton® and Boson®), high-speed event cameras (Prophesee GenX320), servo shields and etc. for robotics, and Ethernet shields for IP camera applications. OpenMV is also working on LTE-M/NB-IoT/GPS cell shields for remote AI applications and a rugged IP67-rated enclosure for outdoor deployments.
Kickstarter Campaign and Availability
Backers can pre-order the OpenMV AE3 for $80, the OpenMV N6 for $120, or both for $200. Additional accessories are available through Kickstarter, with shipping planned before the end of the year. OpenMV has successfully delivered on two previous Kickstarter campaigns, ensuring a proven track record of delivering high-quality embedded vision products.
For more information, visit https://openmv.io or check out our Kickstarter campaign today. Also see below the video of the recent conversation between Kwabena Agyeman, President and Co-Founder of OpenMV, and Jeff Bier, founder of the Edge AI and Vision Alliance and co-founder and President of BDTI.
About OpenMV
Founded in 2015, OpenMV has shipped over 100,000 OpenMV Cams worldwide, enabling AI-powered vision applications in drones, industrial automation, robotics, security, and more. OpenMV’s technology is trusted by industry leaders, including Arduino, and is actively used in cutting-edge AI and IoT projects.
