Nebius Group N.V. is a technology group based in Amsterdam whose main business is providing cloud infrastructure for artificial intelligence. Its offering, presented on nebius.com, is aimed at companies that develop, train or deploy models requiring substantial computing power. Unlike a general-purpose provider covering all computing uses, Nebius focuses on accelerated computing needs, particularly those involving graphics processing units, or GPUs.
From Yandex to a new identity
Nebius has its origins in Yandex N.V., the former Dutch parent company of the group known for its search engine and digital services in Russia. In 2024, the sale of operations in Russia and certain neighboring markets led to a major separation. The Dutch entity retained several international businesses and adopted the name Nebius Group N.V. This break redefined its scope, governance and industrial strategy.
Led by Arkady Volozh, co-founder of Yandex, the group refocused on AI infrastructure. It draws on prior experience in designing distributed systems, servers and data centers. Its return to trading on Nasdaq, under the ticker symbol NBIS, took place in October 2024. The story of Nebius is therefore one of redirecting existing assets and expertise rather than building an entirely new company.
Specialized infrastructure for AI models
The core offering combines GPUs, notably from NVIDIA, with storage resources and networks designed to move data quickly between machines. These components support both model training and inference, meaning the use of models to generate responses or predictions. Nebius develops a software layer for provisioning and managing these resources, with tools tailored to machine learning teams.
Its infrastructure relies in particular on a data center in Finland, complemented by a strategy to expand its capacity in Europe and the United States. The company seeks to control several parts of the technology stack, from server design to cloud operations. For its customers, the aim is to access configurations suited to AI workloads without building their own infrastructure, while retaining visibility into performance and costs.
What comes next?
Nebius’s trajectory will depend on its ability to turn demand for AI computing into a sustainable business. Access to GPUs, power availability, equipment financing and capacity utilization rates are all decisive factors. In competition with major general-purpose cloud providers and other GPU specialists, differentiation will hinge on operational quality, technical support and the actual economics of deployments. Its Dutch base may appeal to European customers, but data localization and protection requirements must be assessed on a service-by-service basis, depending on contracts and hosting regions.