The partnership integrates Nebius's AI-native compute infrastructure directly into Palantir's environment, giving commercial customers tighter control over their data and models.
Palantir Technologies and Nebius Group are joining forces in a deal that positions Nebius as Palantir’s preferred sovereign AI infrastructure partner. The arrangement brings Nebius’s purpose-built AI compute and inference capabilities inside Palantir’s security perimeter, letting eligible commercial customers manage their compute, data, and models without handing the keys to a general-purpose cloud provider.
Both companies trade on the Nasdaq, Palantir under PLTR and Nebius under NBIS, and neither disclosed financial terms.
What the partnership actually does #
At its core, the deal gives Palantir’s commercial customers access to Nebius’s full-stack AI infrastructure without leaving Palantir’s controlled environment. Nebius’s infrastructure was designed from the ground up specifically for AI workloads, which Palantir highlighted as a meaningful differentiator from the generic cloud platforms most enterprises default to.
The partnership also emphasizes open-model deployment alongside proprietary customer data. That means customers can run open-source AI models on their own terms, fine-tuned with their own sensitive information, without that data ever passing through infrastructure they don’t control.
Modular data centers are another piece of the puzzle. Rather than waiting months or years for traditional data center buildouts, the partnership aims to accelerate AI capacity rollout through pre-configured, deployable modules.
The sovereign AI thesis #
Palantir has deep roots in government work, having built its reputation supplying intelligence platforms to defense and security agencies. Nebius, meanwhile, brings AI-native infrastructure expertise that complements Palantir’s software layer. The combination lets both companies pitch a unified stack to commercial enterprises that want government-grade control without actually being a government.
The partnership also aligns with existing collaborative work both companies have done with NVIDIA on operational AI stacks.
Why this matters for the AI infrastructure market #
The emphasis on customer ownership of deployed models is particularly notable. The Palantir-Nebius approach puts the customer in the driver’s seat, with the ability to swap models, adapt to faster solutions, and maintain full oversight of what’s running and where.
For Palantir, this partnership opens commercial revenue streams that don’t depend on government contracts, which have historically been the company’s bread and butter. For Nebius, the deal provides a distribution channel that would be difficult and expensive to build independently, as Palantir already has relationships with large enterprises and government agencies worldwide. No financial terms were disclosed, which leaves the question of economic commitment unanswered. The modular data center component and the integration into Palantir’s security perimeter suggest this sits closer to the substantive end of the spectrum, but execution will determine whether the partnership delivers on its ambitious framing.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our