Inside NEURA’s $1.4 Billion Bet to Build the App Store for Physical AI

Several sources confirmed the complete $1.4 billion sum is contingent on NEURA achieving specific performance milestones, a structure reportedly designed to align long-term interests across such a diverse cap table .

How the $1.4 billion will be used

CEO David Reger has been explicit that this capital is not about modest product iteration. It is infrastructure money, targeted at creating the conditions for a physical AI platform that can support an ecosystem of millions of machines.

NEURA's deployment of the funds prioritizes five core areas :

1. Global production of cognitive and humanoid robots, with an ultimate goal of delivering 5 million robots by 2030 for industrial, service, and household use .
2. Expansion of the Neuraverse platform, an open robotics ecosystem that functions like an operating system and app store for physical machines.
3. Rollout of NEURA Gyms, real-world training centers where hundreds of robots collect multimodal sensor data to build the world's largest repository of physical AI training data .
4. Scaling manufacturing infrastructure, including the company's "NEURA Hive" production method designed for fully automated, high-volume humanoid assembly .
5. Next-generation physical AI systems, deepening sensor fusion, edge inference, and shared learning capabilities across robot fleets.

The scale of these ambitions is reflected in NEURA's reported order book, which already surpasses $1 billion .

Neuraverse: An open platform for physical AI

At the heart of NEURA's strategy lies the Neuraverse, an open platform built on the company's NEURON OS. Rather than treating each robot as a standalone product with proprietary code, the Neuraverse is designed to allow applications to be developed and deployed across different machines—similar to how a smartphone app works regardless of the specific handset brand .

Key architectural principles include:

• Shared real-world learning. When one robot in the network learns a novel task, the capability is distributed across the fleet. This collective intelligence model aims to accelerate capability acquisition exponentially compared to siloed machines .
• Open APIs. The system is designed with application programming interfaces that let third-party developers and industrial partners build on top of the platform, encouraging an ecosystem effect NEURA explicitly compares to the iPhone's impact on smartphones .
• Cloud and edge integration. Through its partnership with AWS, the Neuraverse is hosted on Amazon's cloud infrastructure, leveraging SageMaker for training pipelines that fuse real sensor data with high-fidelity simulation. Meanwhile, Tether's QVAC edge runtime enables local AI inference directly on the robot hardware, reducing latency and cloud dependence .
• Financial layer. Tether's WDK integration creates a native financial layer on the robots themselves, opening potential use cases for autonomous machine-to-machine transactions .

NEURA Gyms: Training in the real world

Simulation has been the dominant paradigm for AI training, but NEURA's leadership argues that robots will only become truly intelligent once they can learn from real-world physical interaction . To that end, NEURA Gyms are large-scale, physical training environments where hundreds of robots collect data simultaneously.

This approach is distinct from purely simulated or text-based training. The robots gather multimodal sensor data—vision, force, torque, proximity—during real manipulation tasks. That data enters the Neuraverse's global repository, becoming a training resource for every connected machine. The company frames this as constructing the world's largest and fastest-growing repository of physical training data, a moat that deepens with every robot deployed .

A particularly ambitious collaboration was announced in March 2026 with the Technical University of Munich (TUM). The partners are establishing a 2,300-square-meter TUM RoboGym at Munich Airport, described as the world's largest robotics research and training center. Hundreds of robot systems, many of them humanoid, will be developed and trained there under the direction of TUM MIRMI professors Lorenzo Masia and Achim Lilienthal .

David Reger's vision for physical AI

Founder and CEO David Reger has become one of the most outspoken advocates for a category he calls "Physical AI"—intelligent systems that move, interact, and learn in the material world rather than being confined to screens and cloud servers.

In the funding announcement, Reger articulated a vision that extends across the entire economy :

"The future of AI will not only live on screens. It will move, interact, learn and work beside us in the real world. We believe Physical AI and cognitive robotics will become one of the largest technology shifts of the coming decades, transforming industries ranging from manufacturing and logistics to healthcare, services and household robotics."

He added a pointed reframing of how the public will evaluate AI progress: "In the future, people will not only ask what AI can say—they will ask what AI can physically do" .

Reger also explicitly positioned the round as evidence that next-generation AI and robotics leadership is not confined to Silicon Valley. A German startup, building hardware and software in parallel, attracting strategic commitments from the world's largest tech corporations, represents a counter-narrative to the assumption that AI's center of gravity is static .

The broader context: A market pivoting from software to hardware

The funding arrives at a moment of structural shift in AI investment patterns. After years of capital pouring into large language models and software-only AI products, the NEURA round signals a widening appetite for "Physical AI"—the category where algorithms meet actuators, sensors, and supply chains. The round valued NEURA at approximately $7 billion, according to the Financial Times, a striking figure for a European hardware company still in the ramp-up phase of commercial deployment .

The presence of Amazon as both an investor and a potential customer is particularly instructive. AWS's role as NEURA's primary cloud provider is a classic enterprise relationship. But the investor release also noted that Amazon will "explore opportunities to deploy NEURA's robotic technologies across select fulfillment center operations" . The line between strategic partner and anchor customer is deliberately blurred.

Similarly, Tether's involvement—including the integration of edge AI runtimes and wallet infrastructure directly into the robotics stack—introduces an economic layer absent from traditional industrial robots. Machines that can transact autonomously may shift the unit economics of deployment in logistics, manufacturing, and eventually services.

Whether NEURA can achieve its 5-million-unit target by 2030 remains an ambitious question. But with $1.4 billion in committed capital, an order book already exceeding $1 billion, and a partner ecosystem spanning silicon, cloud, manufacturing, and finance, the company has assembled the raw components to test whether the Neuraverse can become what it aspires to be: the foundational platform for a generation of machines that don't just analyze the world, but act in it.