Artificial IntelligenceBigTech CompaniesBusinessNewswire

Meta’s $27 Billion Partnership with Nebius: A New Era

▼ Summary

– Meta has signed a new five-year AI infrastructure deal with Nebius Group worth up to $27 billion, significantly expanding their existing partnership.
– The agreement is structured in two parts: $12 billion in dedicated compute capacity and an option for Meta to purchase up to $15 billion in additional capacity.
– The infrastructure will feature one of the first large-scale deployments of Nvidia’s latest Vera Rubin AI accelerator chips.
– Nebius shares surged 14% in pre-market trading on the news, though the financial impact will primarily occur after 2026 as capacity delivery begins in 2027.
– Nebius is a “neocloud” provider, an AI-native cloud company that emerged from the international assets of Yandex after its Russian operations were sold.

A major new alliance is set to reshape the landscape of artificial intelligence infrastructure. Meta has entered into a substantial five-year agreement with the Dutch neocloud operator Nebius Group, a deal valued at up to $27 billion. This partnership, revealed on March 16, 2026, significantly deepens an existing commercial relationship between the two firms. It arrives swiftly on the heels of a separate announcement that Nvidia made a $2 billion strategic investment in Nebius just five days prior. The news propelled Nebius shares upward by 14% in pre-market trading.

The agreement is structured in two distinct phases. The first part involves Nebius providing $12 billion of dedicated computing capacity across several locations, with deployment scheduled to begin in early 2027. Notably, this infrastructure will be constructed around one of the earliest large-scale implementations of Nvidia’s new Vera Rubin platform, the latest generation of AI accelerator chips from the industry leader. The second component grants Meta the option to purchase an additional $15 billion of available capacity from specific upcoming Nebius clusters over the same five-year span. Nebius plans to offer this capacity to other customers first, with any remaining supply then allocated to Meta.

This new contract represents a massive expansion of the companies’ prior engagement. Just months earlier, in November 2025, Meta had signed an initial $3 billion, five-year agreement with Nebius. The latest announcement multiplies that commitment ninefold, bringing Meta’s total contracted spending with the cloud provider to $30 billion. Nebius founder and CEO Arkady Volozh stated the expanded partnership helps secure the large, long-term contracts needed to accelerate the growth of its core AI cloud business. The company noted its 2026 financial guidance remains unchanged, as the capacity delivery and associated revenue will not begin until 2027.

Meta’s enormous financial commitment to AI is a defining corporate strategy. The social media giant has signaled plans to allocate up to $135 billion in capital expenditure toward AI-related projects in 2026 alone, funded largely by its advertising profits. CEO Mark Zuckerberg has also previously outlined an ambition to invest $600 billion in U.S. infrastructure by 2028. This spending frenzy among tech titans is elevating providers like Nebius, which offer specialized, large-scale computing power without the full scope of a traditional hyperscaler, into central roles within the industry’s expansion.

Nebius itself has an unusual history. Headquartered in Amsterdam and listed on Nasdaq, it is the international successor to Yandex NV, once dubbed the “Google of Russia.” Following Russia’s invasion of Ukraine, co-founder Arkady Volozh sold all Russian assets in July 2024 for approximately $2.4 billion, retaining the cloud and data center operations outside the country under the new Nebius brand. Volozh, who was placed on EU sanctions in 2022 and later removed in 2024 after condemning the invasion, now leads a company explicitly positioned as a “neocloud.” This newer class of provider builds data centers from the ground up for GPU-intensive AI workloads rather than adapting general-purpose infrastructure.

The company has been rapidly accumulating major clients. In September 2025, it announced a deal with Microsoft worth an initial $17.4 billion, with potential to reach $19.4 billion. The recent $2 billion strategic investment from Nvidia also included a partnership covering AI factory design, inference stack development, and early access to future Nvidia architectures. Nebius aims to deploy more than five gigawatts of Nvidia systems by 2030 and has existing infrastructure agreements with other large cloud customers.

The deployment of Nvidia’s Vera Rubin platform is a critical aspect of the Meta deal. As the successor to the Blackwell architecture, Vera Rubin’s inclusion in a large-scale commercial rollout signals Nebius’s priority access to cutting-edge hardware, a competitive advantage rooted in Yandex’s historical status as one of Nvidia’s largest customers outside the U.S. and China. It also underscores Meta’s intent to operate at the very frontier of available computing power. Volozh has framed this early access as a structural advantage for building an AI-native cloud. Whether this edge can be maintained as more hyperscalers and neoclouds compete for the same advanced chips remains a key question for the industry’s future.

(Source: The Next Web)

Topics

ai infrastructure 95% corporate partnership 93% nvidia chips 90% financial investment 88% market reaction 85% Cloud Computing 83% compute capacity 82% strategic priority 80% company history 78% geopolitical context 75%