🏗️ AI Infrastructure

Meta Signs Record $27 Billion AI Infrastructure Deal With Nebius — Largest Single Compute Contract in History, First Major Vera Rubin Deployment

2 min read1 views

Meta Platforms has signed a landmark deal with Dutch cloud provider Nebius Group NV worth up to $27 billion over five years for AI infrastructure capacity, making it one of the largest single compute-procurement contracts in history. The deal was announced on Monday, March 16, 2026, and confirmed by both Bloomberg and CNBC.

DEAL STRUCTURE:

  • $12 billion for dedicated AI compute capacity across multiple Nebius locations
  • Up to $15 billion for additional available computing power that Nebius plans to build
  • Total contract value: up to $27 billion over 5 years
  • Nebius will operate one of the first major installations of NVIDIA's latest Vera Rubin AI chips

STRATEGIC CONTEXT:

Nebius founder and CEO Arkady Volozh described the deal as an expansion of existing partnership with Meta, aimed at accelerating growth of Nebius's AI cloud business. Nebius shares jumped 14% in pre-market trading after the announcement.

This deal comes as Meta faces pressure to show returns on its massive AI infrastructure investments. Last November, Meta announced plans to invest up to $600 billion in AI technology, infrastructure, and workforce expansion through 2028. But the high cost is pushing the company to cut back on personnel — reports indicate Meta plans 20% layoffs (approximately 16,000 positions) to fund AI spending.

INDUSTRY IMPLICATIONS:

  1. SCALE OF AI INFRASTRUCTURE RACE: Amazon's CapEx for 2026 is expected to approach $200 billion. The hyperscaler AI spending race is accelerating, with Meta now committing tens of billions to a single cloud provider.

  2. NEOCLOUD EMERGENCE: Nebius joins CoreWeave and Lambda Labs in the 'neocloud' category — specialized AI infrastructure providers that compete with traditional hyperscalers. The $27B Meta deal legitimizes this category at unprecedented scale.

  3. VERA RUBIN EARLY DEPLOYMENT: The deal positions Nebius as one of the first operators of NVIDIA's next-gen Vera Rubin chips, announced at GTC 2026 today. This gives Meta early access to the latest AI accelerators without building all facilities in-house.

  4. GEOGRAPHIC DIVERSIFICATION: Multiple Nebius locations across Europe and potentially Asia provide Meta with geographic diversity for AI workloads, important for latency-sensitive agentic AI applications.

Share this article

🧠 Stay Updated on AI Agents

Get weekly insights on agentic AI, networks and infrastructure. No spam.

Join 500+ AI builders. Unsubscribe anytime.

Deploy Your AI Agent Today

Launch a managed OpenClaw instance in minutes

Request demo →

More from AI Infrastructure