AI everywhere, on every device

Powering AI on billions of edge devices

Home Graphic

Turning hardware
into smartware

Integrate AI software functionality like computer vision, speech recognition or almost any other machine learning functionality into your products, making them smart. Xnor AI models are designed to run in resource constrained environments and are performance optimized on a wide range of diverse hardware platforms.

Run AI locally

  • No data center or cloud connectivity required.
  • Xnor binarized deep learning models are so efficient they can easily run locally, even on battery-operated everyday products.

Minimizes cost

Scales massively

Reduces latency

Safeguards privacy

Powerful
and nimble

Our binary neural networks are up to 10x faster, 30x more power efficient, and need 15x less memory than traditional deep learning models on commodity CPUs. Xnor binarized models run on hardware from the simplest ARM MCUs to the most powerful x86 server CPUs, and can take advantage of performance acceleration hardware like GPUs, DSPs, and custom AI accelerators, including FPGAs.

View Technology

10x Faster


30x More power efficient
15x Less memory

From the creators of YOLO

You Only Look Once

Xnor's award winning technology introduced a novel network architecture designed for the edge. Xnor builds on top of YOLO and other state-of-the-art AI innovations to build custom models of unparalleled efficiency and accuracy, and optimized by being generated using your data and tuned for your hardware.

Xnor's proven AI technology

100+

proprietary
models

1000+

object
categories

5.7

billion
inferences

The Ambarella S5L-0

We're pleased to offer Xnor's advanced AI capabilities on Ambarella's low power, HD camera SoCs. helping to enable a new generation of intelligent IoT cameras, including smart home monitoring devices, capable of delivering analytics at the edge.

Chris Day
VP of Marketing and Business Development Ambarella
Ambarella
GeekWire

"AI for Everyone, Everywhere" may sound like a science-fiction slogan, but it's actually the name given to software from Xnor.ai that's already making devices smarter in the real world…"

Read More >
The New York Times

Ali Farhadi holds a puny $5 computer, called a Raspberry Pi, comfortably in his palm and exults that his team of researchers has managed to squeeze into it a powerful program that can recognize thousands of objects…

Read More >
Financial Times Logo

"XNOR's lightweight algorithms will enable 'AI on the edge' for a broad range of applications, from cars to fleet management to smart infrastructure."

Read More >
A person holds up a mobile device demonstrating XNOR's image recognition software

Join the team putting the power of AI in the hands of everyone

At Xnor we design, build and deliver a range of AI models for applications including object detection and tracking, super resolution, emotion recognition, voice command, and more; we tune each model to deliver optimized performance on any device or platform. Xnor has world-class AI researcher and engineers who are passionate about taking leading-edge innovations from the lab and delivering them as production-ready solutions for real world problems.

Join Us

Want to AI enable your product?

Contact Us