Under the agreement, Nebius will — starting early next year — provide $12 billion of dedicated capacity across multiple locations, the companies announced Monday (March 16).
According to the announcement, the effort will involve “one of the first large-scale deployments of the NVIDIA Vera Rubin platform.” In connection with access to these deployments, Meta has agreed to purchase additional compute capacity for up to $15 billion over a five-year period.
“We are pleased to expand our significant partnership with Meta as part of securing more large, long-term capacity contracts to accelerate the build-out and growth of our core AI cloud business,” Nebius founder and CEO Arkady Volozh said in the announcement. “We will continue to deliver.”
Based in Amsterdam, Nebius rebranded from Russian internet company Yandex. After selling the search engine of that business, it shifted its focus to cloud-computing services for artificial intelligence (AI) operations.
Last month, the company announced plans to acquire agentic search provider Tavily, in a deal aimed at combining real-time search infrastructure with Nebius’s AI platform amid rising enterprise adoption of agentic AI.
We’d love to be your preferred source for news.
Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!
Advertisement: Scroll to Continue
This latest deal is the latest example of Meta’s aggressive spending on AI projects, with the tech giant moving to purchase oChinese startup Manus and, more recently, Moltbook, a social networking platform for artificial intelligence agents.
Aside from its acquisitions, Meta has also been building out its AI talent pool, offering multi-million dollar compensation packages to experts in the field.
The company also plans to spend between $115 billion and $135 billion this year as it races to construct “data centers, chips and infrastructure capable of supporting increasingly powerful AI models,” as PYMNTS wrote last week.
This is a level of spending that places Meta in the same neighborhood as some of the biggest investors in AI infrastructure, including Google, Microsoft and Amazon.
“The difference is that those companies operate cloud businesses that directly monetize the computing power they build,” the report added. “Training models for internal use is only one part of the equation. The same infrastructure can also be rented to thousands of enterprise customers. Meta does not have that outlet.”
Last week saw a report by Reuters that Meta is planning its largest round of layoffs in years, with cuts affecting at least 20% of the company, as it looks to offset its AI spending. A spokesperson for the company dismissed the Reuters report.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.
Facts Only
Nebius will provide $12 billion of dedicated AI capacity to Meta starting early next year.
The deal involves one of the first large-scale deployments of NVIDIA’s Vera Rubin platform.
Meta has agreed to purchase up to $15 billion in additional compute capacity over five years.
Nebius, based in Amsterdam, rebranded from Russian internet company Yandex and now focuses on AI cloud services.
Nebius recently announced plans to acquire agentic search provider Tavily.
Meta has acquired Chinese startup Manus and social networking platform Moltbook.
Meta is offering multi-million dollar compensation packages to AI experts.
Meta plans to spend $115–$135 billion in 2024 on AI infrastructure, including data centers and chips.
Meta does not operate a cloud business to monetize its AI infrastructure, unlike Google, Microsoft, and Amazon.
A Reuters report claimed Meta is planning layoffs affecting at least 20% of its workforce, which the company denied.
Executive Summary
Nebius, a cloud-computing company based in Amsterdam and rebranded from Russian internet firm Yandex, has secured a $12 billion deal with Meta to provide dedicated AI infrastructure capacity starting early next year. The agreement includes one of the first large-scale deployments of NVIDIA’s Vera Rubin platform, with Meta committing an additional $15 billion over five years for compute capacity. Nebius, which recently acquired agentic search provider Tavily, is expanding its AI cloud services amid growing enterprise adoption of AI technologies.
Meta’s aggressive AI investments extend beyond this deal, including acquisitions like Chinese startup Manus and social networking platform Moltbook, as well as high-profile talent recruitment. The company plans to spend $115–$135 billion this year on AI infrastructure, positioning itself alongside Google, Microsoft, and Amazon in AI investment. However, unlike these competitors, Meta lacks a cloud business to monetize its infrastructure, relying instead on internal AI development. Reports of potential layoffs to offset costs have been denied by Meta, though the company’s spending trajectory remains a focal point of industry scrutiny.
Full Take
**Steelman:** This narrative highlights Meta’s aggressive push into AI infrastructure, framing it as a strategic bet on future dominance in AI-driven technologies. The partnership with Nebius underscores the scale of investment required to compete in AI, while the lack of a cloud monetization outlet distinguishes Meta’s approach from its peers. The denial of layoff reports suggests confidence in long-term growth despite short-term costs.
**Pattern Scan:** The article leans into a "big tech arms race" framing, emphasizing spending figures and acquisitions to create a sense of urgency and inevitability around AI dominance. The contrast between Meta’s spending and its lack of cloud revenue could subtly imply recklessness, though the piece avoids overt judgment. The inclusion of a denied layoff report introduces a tension between ambition and sustainability, a common narrative device in tech coverage.
**Root Cause:** The underlying paradigm is the belief that AI infrastructure is a zero-sum game—whoever builds the most capacity fastest will control the future. This assumes that scale alone guarantees success, ignoring potential inefficiencies or alternative paths to AI leadership. The unstated assumption is that Meta’s spending is justified by future returns, though the article notes the lack of direct monetization avenues.
**Implications:** For human agency, this raises questions about concentration of power—fewer entities controlling AI infrastructure could limit innovation and autonomy. The costs are borne by Meta’s shareholders and employees (if layoffs materialize), while benefits accrue to executives and investors betting on long-term AI dominance. Second-order effects may include accelerated AI development but also potential market distortions if smaller players are priced out.
**Bridge Questions:**
How might Meta’s lack of a cloud business affect its ability to sustain this level of investment compared to competitors?
What alternative strategies could yield AI leadership without such massive capital expenditure?
If layoffs do occur, how does that reconcile with the narrative of AI-driven growth?
**Counterstrike Scan:** A coordinated influence campaign might exaggerate Meta’s spending to either hype AI as an unstoppable force (benefiting investors) or critique it as reckless (benefiting competitors or regulators). The actual content presents facts without overt manipulation, though the framing of "aggressive spending" and "largest layoffs" could subtly nudge readers toward skepticism. No clear alignment with a hypothetical attack playbook is detected.
Patterns detected: none
Sentinel — Human
The article shows strong signs of human authorship, with natural stylistic variations and specific sourcing, though minor template-like elements suggest possible editorial standardization.