Hadrian raises $1.37B as factories-as-a-service scales to Navy, Army, Lockheed, and international defense

Hadrian raises $1.37B as factories-as-a-service scales to defense and beyond

Hadrian has closed $1.37 billion in new funding, its largest raise to date, as the company scales what it calls factories-as-a-service across the U.S. defense industrial base. CEO Chris Power says the round reflects both the adoption of that model at mass scale and a regulatory environment that is finally moving in Hadrian's favor, including the recent ban on Chinese components in missiles and drones.

What Hadrian actually sells

The core product is a design-agnostic, highly automated factory that Hadrian builds and operates on a customer's behalf. Customers bring the IP; Hadrian handles site selection, development, construction, and a decade's worth of operations. Power describes the revenue model as resembling CoreWeave plus Palantir — factories-as-a-service as the base layer, with Opus, Hadrian's physical AI and factory autonomy platform, sold on top. Three enterprise customers have already licensed Opus directly and are deeply integrated with it.

Current customers include the U.S. Navy, Army, Lockheed Martin, and other defense primes. Power says the company is growing toward billions in revenue.

“Raw scaling and then the proof that factories as a service was a desperately needed to create and that people are adopting it at mass scale across the Navy, the Army, our customers like Lockheed Martin... We have LA, Arizona and Alabama is under construction — 3,000,000 square foot in total.”

Underwriting and utilization

Working with startups directly is largely off the table for now, but Hadrian is building partnerships with smaller industrials companies and integrating their capabilities into its factories-as-a-service stack, effectively bringing them up into the enterprise supply chain rather than serving them as end customers.

Underwriting manufacturing risk is easier with primes because of their revenue predictability. Hadrian's structural advantage is that at least 80% of its CapEx can be redeployed if a single program disappears, since its factories operate as virtual, flexible capacity rather than single-line production facilities.

Footprint

Hadrian is currently operational in Los Angeles and Arizona. An Alabama facility is under construction, bringing total planned production space to 3 million square feet. On the R&D and physical AI side, the company has 500,000 square feet in LA, with another 1 million square feet under construction there for software engineering and physical AI testing. A San Francisco location will be announced within the next month.

Where growth comes from next

Power is direct that any market pivot right now would be doing a disservice to the mission. Demand across submarines, munitions, drones, energetics, and the broader Army and Navy organic industrial base is absorbing all available capacity, and international allies seeking defense re-industrialization are adding to that pipeline.

Commercial manufacturing — particularly robotics supply chains, as the U.S. moves to ban Chinese consumer electronics components — is the next candidate for Hadrianification. But Power's view is that the sequencing matters: build defense capacity first, then let commercial demand fill in around it.

Technology adoption and qualification bottlenecks

One underappreciated constraint Power raises is the qualification cycle for new manufacturing methods. Getting a novel welding or additive technique approved for defense and aerospace use takes roughly a year and 30 engineers just to submit an engineering record. Hadrian sees an opportunity to accelerate adoption for smaller companies that can't eat that cycle alone, using its factory partnerships as the qualification pathway for new casting, additive, and laser welding techniques that exist but haven't yet scaled.

Power also hinted, without confirming, that Hadrian may be developing its own physical AI models — stopping himself mid-sentence when he thought something had been leaked.