NextFin News - Nebius Group NV has secured a combined $46 billion in artificial intelligence infrastructure commitments from Meta Platforms and Microsoft, a staggering sum that effectively triples the company’s current market capitalization in contracted value. The deals, finalized in early April 2026, position the European-based cloud provider as a critical third-party "AI factory" for the world’s largest technology firms as they race to secure the specialized compute capacity required for frontier model training.
The centerpiece of this expansion is a five-year agreement with Meta Platforms, which has committed to spending up to $27 billion on Nebius’s Nvidia-powered infrastructure. According to terms disclosed in the deal, Meta will take $12 billion of dedicated capacity starting in 2027, with an option for an additional $15 billion in on-demand capacity. This follows a separate $17.4 billion contract with Microsoft, which had already validated Nebius’s technical stack earlier in the year. Together, these agreements represent one of the largest private-sector infrastructure build-outs in the history of the cloud industry.
For Nebius, which currently carries a market valuation of approximately $28 billion, the $46 billion in total deal value creates a rare disconnect between contracted revenue and equity pricing. The company, led by Arkady Volozh, has pivoted aggressively toward high-performance computing after spinning off its Russian assets. This transformation has turned a former search engine operator into a pure-play AI infrastructure provider, operating massive data centers in Finland and expanding across Europe and the United States to meet the insatiable demand for H100 and Blackwell-class GPU clusters.
The scale of these commitments reflects a shift in how "Big Tech" manages its capital expenditure. Rather than building every data center in-house, companies like Meta and Microsoft are increasingly outsourcing the specialized, high-density power requirements of AI training to nimble providers like Nebius. This strategy allows the giants to scale faster than their own construction timelines permit, while Nebius benefits from the long-term revenue certainty required to finance its own multi-billion dollar hardware purchases from Nvidia.
However, the market’s reaction remains tempered by significant execution risks. Analysts at Simply Wall St have noted that Nebius’s recent share price volatility and a high level of non-cash earnings could complicate the interpretation of its reported profitability. The firm’s reliance on customer financing and upfront payments to fund its massive hardware orders introduces a layer of liquidity risk if deployment timelines slip or if the cost of power and cooling exceeds initial projections. Furthermore, the concentration of revenue among just two or three massive clients creates a "key customer" risk that few other cloud providers face at this scale.
The geopolitical context also looms over Nebius’s rapid ascent. As U.S. President Trump’s administration continues to emphasize domestic technological supremacy and tightens controls on AI hardware exports, Nebius must navigate a complex regulatory landscape. While its European headquarters provides some insulation, its heavy reliance on American-made chips and American-headquartered customers means its operations are inextricably linked to Washington’s trade policies. Any shift in export licenses or data sovereignty laws could disrupt the delivery of the very capacity Meta and Microsoft have just purchased.
Despite these hurdles, the sheer volume of the contracts suggests that the "compute crunch" is far from over. By locking in $46 billion in deals, Nebius has effectively cornered a significant portion of the available high-end GPU capacity for the next half-decade. The success of this venture now rests on the company’s ability to turn these massive paper commitments into operational data centers without succumbing to the inflationary pressures of the global energy market.
Explore more exclusive insights at nextfin.ai.
