Stereolabs is powering the next generation of humanoid robots

The age of humanoids and physical AI is no longer a vision of the future — it’s happening now. Humanoid robots are stepping out of research labs and into the real world: factories, warehouses, offices, and even homes.

But to operate in our world, these robots need more than just actuators and code — they need human-like vision.

This is where Stereolabs leads, delivering end-to-end perception systems that equip humanoids with the spatial awareness, localization, and scene understanding required to navigate complex, dynamic environments in real time

Meet next generation humanoids powered by ZED

Leading companies across the humanoid robotics ecosystem are integrating Stereolabs vision systems as their perception backbone. From startups to enterprise platforms, ZED cameras and SDK are quickly becoming the perception engine of choice for next-generation humanoids and physical AI.

Here are some of the humanoids already leveraging our technology:

4NE1by Neura Robotics

New generation of 4NE1 using ZED Cameras

MenteeBot by Mentee Robotics

Westwood Robotics’

DROID project with 15 top universities

These platforms represent just a few examples of how Stereolabs and ZED are helping redefine what humanoids can perceive and do. Many other companies are currently integrating ZED in their upcoming design. Whether it’s for mobility, manipulation, or interaction, ZED cameras are the perception engine behind the next wave of physical AI.

End-to-end perception systems for humanoids

Why are humanoids companies choosing Stereolabs for their vision provider? At the core of every next-gen humanoid is the ability to perceive, localize, and manipulate — all in real time. Stereolabs provides a complete perception and localization stack that combines both stereo and mono cameras with the industry’s most complete vision SDK and ROS integrations.

• ZED X Mini, a compact stereo camera delivering high-resolution 3D spatial data, perfect for depth sensing, obstacle detection, and fine-grained manipulation.
• ZED X One, a high-performance global shutter 2D camera built for visual tracking, recognition, and stereo pairing, allowing maximum flexibility across robotic form factors.
• GMSL capture cards, like the ZED Link Quad, that supports up to 8 2D and 3D cameras on platforms such as Nvidia AGX Orin and AGX Thor — enabling scalable multi-camera systems with ultra-low latency.
• ZED Box Mini, a new, highly efficient design, delivering the same performance in a more compact and cost-effective form. Supports up to 2 GMSL2 cameras, making it ideal for large-scale deployments.

Together, our vision hardware and software stack enable robots not only to "see" but to understand their environment, plan their motion, and interact with the physical world — just like humans do.

The future of physical AI is here

At Stereolabs, we believe that physical AI is the next great technological leap — and vision is its most important sense.

The adoption of ZED vision systems by cutting-edge humanoid platforms is validation of a decade-long focus on performance, reliability, and developer experience. Our technology doesn’t just help robots see—it helps them thrive.

Whether you’re building the next Atlas or logistics robot, we’re here to help bring your vision to life.

Start building with ZED today

Physical AI needs eyes — and ZED delivers. Stereolabs stereo and mono camera systems provide the critical perception backbone for robots to see, localize, and interact with the world around them.

Whether you're building mobility, precise manipulation, or full-scene understanding, order your physical AI-ready vision system with ZED X and ZED X One and fast-track your next-gen robotics deployment.