Nebius Group N.V. (NBIS) saw its shares climb sharply on Monday after announcing a $643 million acquisition of U.S.-based Eigen AI, a move that deepens the Amsterdam-based company's push into the competitive AI cloud market. The deal, disclosed in a May 1 SEC filing, combines cash and stock to bring Eigen's model optimization technology in-house.
Shares of Nebius closed at $154.49 on Friday, marking an 11.76% gain, and premarket trading on Monday indicated further upside. The acquisition is structured as a cash-and-stock transaction, with up to $98 million in cash and approximately 3.8 million Nebius Class A shares, subject to final terms. Notably, the deal does not require shareholder approval, allowing for a streamlined integration process.
Eigen AI, which operates as MagicByte Inc., specializes in inference and model optimization—a critical area where AI models generate responses after training. By folding Eigen's technology into its Nebius Token Factory managed inference platform, Nebius aims to offer developers a more efficient and cost-effective path to production AI. Even small gains in computing efficiency can yield significant savings at scale, a key selling point for enterprise clients.
The acquisition also establishes a new engineering and research hub in the San Francisco Bay Area, led by Eigen's founders. To retain key talent, the deal includes a four-year vesting schedule for shares, with founders and continuing employees receiving only 15% upfront. This structure is designed to align incentives and prevent attrition during the integration period.
Roman Chernin, co-founder and chief business officer at Nebius, emphasized the need for 'optimized inference and infrastructure scale' in the current AI landscape. Eigen AI's co-founder and CEO Ryan Hanrui Wang echoed that sentiment, stating the goal is to let developers 'run models reliably in production' without worrying about the underlying infrastructure.
This acquisition places Nebius squarely in the AI compute arms race, competing with heavyweights like CoreWeave and other niche cloud providers. The company is betting on a strategy that combines Nvidia-backed hardware with its own proprietary software stack. Reuters has labeled Nebius a leading 'neocloud' company, renting out hardware and cloud services to tech firms focused on AI workloads.
The timing is tight for investors. CoreWeave is set to report first-quarter results on May 7, offering an early gauge of demand, margins, and costs for AI infrastructure players. Nebius follows on May 13, when its own numbers will reveal whether capital spending aligns with actual demand—a more critical metric than any deal headline.
Nebius has been aggressively expanding its client base and supplier partnerships. In March, the company announced a five-year AI infrastructure deal with Meta that could total around $27 billion, with $12 billion locked in for dedicated capacity and an option for Meta to purchase additional compute. Earlier that month, Nvidia committed $2 billion to Nebius to support a rollout of more than 5 gigawatts of Nvidia systems by 2030. That investment followed a multi-year infrastructure deal with Microsoft, centered on reserved capacity from a Vineland, New Jersey data center.
Despite the momentum, the Eigen AI deal is not without risks. The SEC filing notes that closing depends on standard conditions, including antitrust clearance. Nebius also cautioned that integration challenges, client retention, and securing funding could weigh on performance. For a company racing to expand data center capacity amid elevated AI infrastructure costs, these factors are far from trivial.
For now, investors are treating Eigen AI as an additional layer on top of Nebius's broader infrastructure push. The real test comes on May 13, when the company reports earnings—where the focus will shift from deal headlines to whether spending is translating into sustainable revenue growth.
