In just a matter of weeks in early 2026, a relatively under-the-radar Dutch company named Nebius Group (NASDAQ: NBIS) has emerged as arguably the most important AI backend infrastructure partner on the planet. With mega-deals signed with Meta, Microsoft, and NVIDIA in rapid succession, Nebius is rewriting the rules of how the world’s biggest tech companies secure their AI compute power.
What Is Nebius?
Nebius is an AI cloud company building a full-stack platform for developers and companies — covering everything from data and model training to production deployment. Founded on deep in-house technological expertise and operating at scale with a rapidly expanding global footprint, Nebius serves startups and enterprises building AI products, agents, and services worldwide. It is listed on Nasdaq (NBIS) and headquartered in Amsterdam.
Nebius was founded in 2022 after a restructuring of Russian company Yandex’s operations based outside of its home market and listed in New York in 2024. What makes its rise so remarkable is the sheer speed at which it has gone from a restructured tech spin-off to a mission-critical backbone for AI’s biggest players.
The NVIDIA Partnership: $2 Billion and a 5 Gigawatt Vision
The turning point happened on March 11, 2026. That’s when NVIDIA and Nebius revealed a major infrastructure deal in AI history.
NVIDIA will invest $2 billion in Nebius. NVIDIA trusts Nebius’s business and its strong AI engineering skills. The partnership deepens their relationship across the full stack — from AI factory architecture to production software — enabling Nebius to accelerate the buildout of its industry-leading, full-stack AI cloud platform.
To enable Nebius to deploy more than 5 gigawatts of capacity by the end of 2030, the companies will collaborate on AI factory design and support, inference stacks using NVIDIA’s latest software technologies and optimised models, and AI infrastructure deployment — including early adoption of the NVIDIA Rubin platform, NVIDIA Vera CPUs, and NVIDIA BlueField storage systems.
This isn’t just a hardware procurement deal — it positions Nebius as a preferred engineering partner with privileged access to NVIDIA’s next-generation chips before they reach the broader market.
The Meta Deal: $27 Billion and One of History’s Largest Compute Contracts
Just five days after the NVIDIA announcement, Nebius dropped another bombshell.
Under a five-year agreement announced on March 16, 2026, Nebius will provide $12 billion of dedicated capacity across multiple locations, based on one of the first large-scale deployments of the NVIDIA Vera Rubin platform. Meta will buy more compute capacity in the upcoming Nebius clusters. They plan to spend up to $15 billion over five years.
The deal is one of the largest compute-procurement contracts ever. It marks a clear shift in the global AI landscape. This also shows a bigger trend. Hyperscalers aren’t just relying on their own data centres anymore. They are also moving away from the “Big Three” cloud providers. They are now looking to specialised firms like Nebius for large GPU clusters.
The Microsoft Deal: Up to $19.4 Billion Over Five Years
The Meta partnership wasn’t even Nebius’s first mega-deal of this cycle. In September 2025, Nebius inked a deal to deliver computing resources to Microsoft, worth up to $19.4 billion over five years. That contract helped set the stage for the company’s exponential stock performance — Nebius’s share price rose more than 200% in 2025 and has increased by 35% so far in 2026.
What Makes Nebius Different? The “Token Factory” Model
The establishment of Nebius Token Factory represents the company’s most strategically significant product decision to date. Rather than selling pure compute, Token Factory delivers managed inference endpoints where the SLA is defined by token throughput, latency, and cost. This distinction is important because AI economics are changing from training to inference.
Nebius holds a “Preferred Partner” status with NVIDIA. This lets them access advanced chips early. This creates a strong advantage that’s tough for newcomers to match. Nebius has liquid-cooling innovations and unique data centre designs. They enjoy gross margins of about 40%. This is much higher than typical hardware providers.
Why This Matters for the AI Ecosystem
Meta is part of a group of hyperscalers planning huge spending as they race to build AI infrastructure, with a combined $700 billion in spending planned by hyperscalers, including Amazon, Alphabet, and Microsoft. The fact that all of these giants are looking beyond their own walls for compute — and finding Nebius — tells you everything about where AI infrastructure is heading.
Nebius is no longer just an alternative cloud provider. It is fast becoming the default AI backend for the companies building the future.
Recommended
