Nebius: Capital Raising

The leadership read

Nebius entered 2025 as a company in transition, spun out of Yandex's restructuring and carrying legacy complexity, and the $1B raise committed it to a materially different posture: purpose-built AI cloud infrastructure competing directly with hyperscalers for the GPU cluster workloads that European AI labs and enterprises are actively moving off commodity cloud. That shift means Nebius is no longer a repackaged cloud reseller; it has now staked its capital structure on owning the full stack from silicon access to managed inference, a bet that demands delivery at datacenter scale, not just platform iteration. This is one of twelve capital-raising signals in AI and compute infrastructure we have tracked in the last 90 days. The relevant comparables are not the seed rounds in the set, Intropy, AI Infrastructure Capital AG, but the scale transactions: Nscale's $1.53B across two 2025 rounds and Stack's A$5.9B datacenter financing facility in Australia. The consistent shape across these deals is capital concentration in physical and GPU-dense infrastructure, with software and managed services layered on top to defend margin. The pattern is accelerating, not plateauing. Companies reaching this stage of capital density in AI infrastructure consistently face rising demand for three functional areas: datacenter operations leadership with experience delivering at sovereign or regulated-market scale, commercial leadership capable of building enterprise and government sales motions distinct from developer self-serve, and technical product leadership at the seam between hardware procurement and managed-inference product design. The talent pool with all three is thin in EMEA.