Neuro N6
Arduino-compatible edge AI development board. Up to 600 GOPS of neural compute under 750mW.
Pre-order nowFrom £89 · Ships late 2026
Neuro N6 pre-orders live · 500%+ funded on Kickstarter
Edge AI, made practical.
Auto dataset labelling, augmentation, and export for vision AI. Runs free in the browser.
Try Pixel Kit freeNo signup required
Featured on CNX Software500%+ funded on KickstarterFeatured on Hackster.io174 backersFeatured on NotebookcheckFunded in under 4 hoursShipping November 2026
0 GOPSPeak NPU compute
0msYOLOv8n inference
<0mWTotal power draw
0MPISP camera support
The Platform
Everything between idea
and deployment.
Neuro N6
Arduino-compatible edge AI development board. Up to 600 GOPS of neural compute under 750mW, in a Feather form factor.
STM32N6Cortex-M55Neural-ART NPUUSB-C
Vision AI without the toolchain. Auto-label, augment, and export datasets to YOLO, COCO, or Pascal VOC.
Free tier · No signupTry free
Neuro Studio
Live detections, confidence scores, and timing streamed over USB. Ships with the board.
Neuro Vision modules
Four swappable camera modules: 5MP autofocus, 237 FPS global shutter, radiometric thermal, and wireless.
From £29Compare
Open source
The Arduino core and hardware design files are on GitHub, with more landing as development progresses.
github.com/ohmlab-ltdFor industry
Applied AI and end-to-end consultancy: custom hardware, embedded software, and on-device AI deployment.
ExploreThe Workflow
Dataset to deployment, one pipeline.
01
Capture
Swappable Neuro Vision modules: rolling shutter, global shutter, or radiometric thermal, straight onto the board.
02
Label and train
Pixel Kit auto-labels your images, augments them, and exports a training-ready dataset in YOLO, COCO, or Pascal VOC.
03
Deploy
The whole inference pipeline lives in one Arduino sketch. If you can blink an LED, you can run a neural network.
04
Observe
Neuro Studio streams live detections, confidence scores, and timing over USB. No config required.
For IndustryIn development
Vision AI where it matters most.
Active discovery across multiple sectors, identifying where applied AI can deliver the highest real-world value. Meaningful problems, evidence-led solutions, reliable performance in operational environments.
- Custom hardware design, from concept to production-ready PCB
- Embedded and edge software, built to be reliable and updatable
- AI model deployment, from training to a measurable on-device system
Build at the edge.
Whether you're prototyping on a bench or exploring a use case for your operation, we'd love to hear from you.