Nebius has just secured one of the more strategically important endorsements available in enterprise artificial intelligence. Palantir Technologies has named Nebius its preferred sovereign AI infrastructure partner, with the two companies planning to combine Palantir’s enterprise software and data architecture with Nebius’s AI-native cloud, compute capacity and inference infrastructure. The partnership, announced before the U.S. market opened on Tuesday, September 8, is designed to let eligible Palantir commercial customers run open AI models on Nebius infrastructure while retaining control over their own data, models and computing environment. No contract value, minimum revenue commitment or guaranteed volume was disclosed, which means investors should treat the announcement as a strategic distribution agreement rather than a multibillion-dollar customer contract—at least for now.
That distinction is crucial for both Nebius stock and Palantir stock. Palantir is effectively bringing a new infrastructure layer inside its enterprise perimeter, while Nebius gets privileged access to one of the fastest-growing customer bases in enterprise AI. The companies also intend to work together to bring additional computing capacity online faster, including through modular data-center deployments at locations where electricity is already available. In an AI market where access to GPUs, power and secure infrastructure can delay deployments by months or even years, that second part of the agreement may ultimately prove as important as the software integration itself.
The partnership therefore attacks one of the biggest bottlenecks in enterprise AI from both directions. Palantir provides the operating environment that allows companies to connect models with proprietary data and real-world workflows. Nebius supplies the specialized infrastructure required to train and operate those models. If the combination works as intended, customers could move from AI experimentation to production without surrendering their most valuable data to a third-party model provider. That is exactly the kind of sovereignty argument Palantir CEO Alex Karp has been pushing aggressively—and the latest growth numbers suggest customers are listening.
Palantir Is Giving Nebius Something More Valuable Than a Normal Cloud Partnership
The most important phrase in Tuesday’s announcement is not simply “strategic partnership.” It is “preferred sovereign AI infrastructure partner.” Palantir said that, following an integration period, Nebius compute and inference endpoints will be brought inside the Palantir enterprise perimeter. Eligible customers will then be able to access Nebius infrastructure through Palantir’s Sovereign AI stack while keeping control of their compute, data and models.
Palantir describes its Sovereign AI Operating System as being built around AIP, Foundry, Apollo and Ontology, with those products acting as the authorization, operational and isolation layers between sensitive enterprise data and the models being deployed against it. The Nebius partnership adds the physical and cloud-computing layer underneath that architecture. Instead of a company sending proprietary information into a closed external model and hoping governance controls are sufficient, the vision is that customers can run open models on infrastructure they trust and continuously adapt those models using their own domain-specific data.
Related articles
For Palantir, that strengthens an increasingly important sales pitch. For Nebius, however, the commercial implications may be even more interesting because Palantir already has access to large enterprises and government-adjacent customers that smaller AI infrastructure providers can struggle to reach. Nebius does not need to build every customer relationship from scratch if Palantir can help pull its infrastructure into deployments where AIP and Foundry are already embedded.
That does not mean revenue will appear automatically. The press release explicitly identifies risks around successful joint marketing, customer adoption and Nebius’s ability to expand revenue through its preferred-partner status. No purchasing obligation was announced. But becoming the preferred infrastructure layer inside a platform growing as quickly as Palantir gives Nebius a potentially valuable new distribution channel.
Palantir’s Explosive Growth Makes That Distribution Channel Hard to Ignore
The timing could hardly be better for Nebius. Palantir reported $1.94 billion of second-quarter revenue, up 93% from a year earlier, while U.S. commercial revenue surged 149% to $764 million. U.S. government revenue rose 90% to $809 million, and the company closed 220 deals worth at least $1 million during the quarter, including 73 contracts worth $10 million or more. U.S. commercial remaining deal value reached $6.24 billion, representing 124% year-over-year growth.
Those figures turn the Nebius agreement into more than a technology demonstration. Palantir is currently adding enterprise AI workloads at an extraordinary pace, while management has raised full-year 2026 revenue guidance to approximately $8.15 billion and expects U.S. commercial revenue to exceed $3.42 billion. Karp has repeatedly framed that growth around AI sovereignty: customers want the benefits of advanced models without handing control of proprietary operational data to external providers.
Nebius now gets positioned directly underneath that narrative. If even a portion of new Palantir commercial deployments require dedicated AI compute, managed inference or sovereign infrastructure, Nebius could participate in spending that previously might have flowed to the largest hyperscale clouds or remained on-premises. The partnership could become especially relevant for regulated industries, strategically sensitive enterprises and organizations that want access to open models without locking themselves into a single proprietary AI provider.
Why Nebius Fits Palantir’s Sovereign AI Pitch Almost Perfectly
Nebius has been building precisely the type of infrastructure Palantir is looking for. Rather than operating as a conventional general-purpose cloud that later added GPU services, Nebius describes its platform as designed specifically for AI—from hardware architecture and networking through storage, orchestration, training and inference. Palantir explicitly cited that AI-native design as one reason it selected the company.
The infrastructure story has also become much larger during 2026. Nvidia announced in March that it would invest $2 billion in Nebius as part of a strategic partnership intended to help Nebius deploy more than 5 gigawatts of Nvidia systems by the end of 2030. The companies are collaborating on AI-factory architecture, inference and agentic AI, giving Nebius unusually close ties to the dominant supplier of accelerated-computing hardware.
Nebius has simultaneously demonstrated that major technology companies are willing to commit enormous sums to its capacity. In March, the company announced a five-year agreement under which Meta will receive $12 billion of dedicated capacity based on Nvidia Vera Rubin systems, while Meta can purchase additional available capacity worth up to $15 billion over the same period under the broader arrangement.
Palantir brings something different from Meta. Meta validates Nebius as a hyperscale infrastructure provider capable of handling enormous workloads. Palantir potentially validates Nebius as an enterprise infrastructure platform that can sit underneath many customers rather than relying solely on a handful of gigantic technology buyers. For NBIS investors, diversification of that customer base could eventually matter almost as much as raw contract value.
The Real Opportunity Is Inference, Not Just Training
Another important detail is Palantir’s decision to integrate Nebius inference endpoints, not merely rent GPU clusters for model training. Training frontier AI models generates spectacular capital-expenditure headlines, but inference—the computing required every time a deployed model answers a query, analyzes a document or executes an agentic workflow—can create more persistent utilization if enterprise AI adoption scales.
Nebius has been building that part of the stack through its Token Factory platform and partnerships around optimized open-source models. The company has argued that organizations increasingly want customizable models that can be operated against private data without being tied permanently to proprietary APIs. That vision overlaps almost perfectly with Palantir’s argument that customers should retain ownership and control over the intelligence generated from their own operational information.
The two companies are effectively betting that the next phase of AI will be less about every enterprise sending requests to the same handful of closed frontier models and more about organizations operating specialized models tuned to their own industries and workflows. If that happens, Palantir can orchestrate the data and business logic while Nebius provides the underlying computing capacity. The better those specialized models become, the more inference workloads Nebius could potentially serve.
Nebius Is Growing Fast Enough That the Partnership Could Actually Matter
Nebius is no longer operating at startup scale. Its second-quarter investor materials showed group revenue of approximately $582 million, up 454% year over year, while the core Nebius AI Cloud business reached an adjusted EBITDA margin around 50%. The company ended the quarter with roughly $8 billion of cash and highlighted more than 4 gigawatts of contracted power capacity as it races to bring additional data centers online.
That expansion requires enormous amounts of capital. In August, Nebius completed a $5.75 billion convertible-note offering, saying proceeds would be used for data-center construction, GPU purchases, expansion of its AI cloud and other growth investments. The financing underscores both sides of the NBIS investment case: demand is large enough for management to justify aggressive expansion, but building the infrastructure necessary to capture that demand is extremely capital-intensive.
The Palantir partnership could help reduce one of the biggest risks associated with that spending—finding enough high-quality customers to utilize the capacity once it becomes available. Nebius has already developed a model in which outside infrastructure partners can finance and own data centers while Nebius provides the architecture, software stack and go-to-market platform, potentially allowing the company to scale with less incremental capital. Palantir’s announcement specifically mentions modular deployments at sites where power is already available, suggesting the companies are thinking not only about selling existing capacity but also about expanding it around customer demand.
Why the Deal May Be More Important for NBIS Stock Than PLTR Stock
For Palantir shareholders, the partnership strengthens the Sovereign AI product offering but does not fundamentally alter a company already approaching $2 billion of quarterly revenue with enormous margins and billions in contracted business. Palantir can potentially offer customers more infrastructure flexibility without having to finance a global GPU cloud itself.
For NBIS stock, the signal is potentially larger. Nebius needs to prove that its rapidly expanding infrastructure can win high-value enterprise workloads beyond its giant hyperscaler agreements. Being selected as Palantir’s preferred sovereign infrastructure partner provides third-party validation of Nebius’s architecture and puts the company in front of customers already spending aggressively on operational AI.
There is still no disclosed financial commitment, so assigning billions of dollars of future revenue to the partnership would be speculation. That is the most important restraint for investors evaluating Tuesday’s announcement. The deal becomes financially meaningful only when Palantir customers begin consuming Nebius capacity at scale.
But that is also what makes the next phase worth watching.
Palantir and Nebius May Be Building the Missing Middle of Enterprise AI
The AI market is increasingly splitting into several layers. Nvidia supplies the accelerated computing hardware. AI laboratories create increasingly powerful models. Hyperscalers operate enormous general-purpose clouds. Palantir specializes in connecting AI with an organization’s actual data, permissions and workflows. Nebius is attempting to occupy the infrastructure layer between those worlds: specialized AI compute with the flexibility to train, customize and serve open models.
Tuesday’s partnership connects those layers in a way that could become particularly attractive as governments and corporations focus more intensely on digital sovereignty. A recent Capgemini survey reported by Reuters found that large companies and public organizations are increasingly treating digital infrastructure as a strategic dependency comparable with physical supply chains and energy, with executives emphasizing the importance of retaining control over critical data and AI models.
That backdrop makes the timing notable. Palantir’s software says customers should control their data and intelligence. Nebius says customers should control the infrastructure and models running beneath it. Together, they are making the case that enterprises do not have to choose between cutting-edge AI performance and sovereignty.
For Nebius stock, the partnership is therefore less about an immediate revenue number and more about validation. Nvidia has already backed Nebius technologically and financially. Meta has validated its ability to deliver enormous AI infrastructure contracts. Palantir is now giving Nebius a route into enterprise sovereign AI.
The next milestone will determine whether Tuesday’s announcement deserves to become a major part of the NBIS investment thesis: actual Palantir customers deploying models, consuming Nebius inference capacity and turning “preferred partner” status into recurring revenue.
If that starts happening at scale, this partnership could look considerably more important a year from now than it does on announcement day.
Disclaimer
This article is for informational purposes only and does not constitute financial or investment advice. Readers should conduct their own research and consider consulting a qualified financial advisor before making investment decisions. This article was researched and drafted with the support of AI, then reviewed, fact-checked and edited by the editorial team before publication.
