Nebius Group acquires Inferize to bolster AI inference platform

Nebius Group, the Amsterdam-based AI cloud provider, has acquired Inferize, a startup specializing in AI inference optimization. The acquisition aims to integrate Inferize’s technology into Nebius’s enterprise inference platform, Token Factory.

The move is designed to reduce ‘cold start’ delays—the time required to load large AI models—thereby minimizing GPU idle time and allowing compute capacity to scale more efficiently with actual demand. While the specific financial terms of the deal were not disclosed, the integration follows previous technical expansions of Token Factory through acquisitions like Eigen AI and Clarifai.

Nebius continues to see significant growth in the neocloud sector, reporting US$575 million in AI cloud revenue for the second quarter of 2026. The company has secured major contracts, including a multi-billion dollar agreement with Meta and a partnership with Palantir. Nebius is also expanding its physical footprint, aiming to have 16 data centers operational by the end of the year.