Anthropic Signs a $35 Billion Deal With Lambda to Lock In Compute Before the Shortage

By Carlos Montiel | Enterprise AI Specialist
Leer en español →
Published: 2026-09-05 | By: Carlos Montiel | Reading time: ~4 minutes

Anthropic has signed a deal to rent $35 billion worth of compute capacity from cloud provider Lambda, backed by Nvidia, as part of an increasingly frantic race to lock in GPU capacity ahead of a shortage the industry itself is anticipating. With this deal, Anthropic now has $175 billion in cloud commitments signed in just the past few months.

A three-way deal with Nvidia at the center

According to reports from Bloomberg, The Information, and The Wall Street Journal, the deal's structure is unusual: Anthropic pays Lambda, Lambda installs Nvidia hardware, and Nvidia pays the rent on the underlying facility to Hut 8, which hosts the data center. Nvidia reportedly holds the lease on the data center itself, making it the central piece of the transaction even though the primary contract is between Anthropic and Lambda.

Where and how much capacity

The data center involved is being developed by Hut 8 at its Beacon Point campus in Nueces County, Texas, with roughly 350 megawatts of capacity. It's one of several pieces of physical infrastructure Anthropic is locking in ahead of time for its Claude models, at a moment when training and inference compute demand keeps outpacing the available supply of high-end GPUs.

The number that matters: $175 billion in cloud commitments racked up in a matter of months isn't normal operating spend — it's a bet that compute capacity will be the competitive bottleneck of the next few years, and that locking it in now, even at enormous cost, is worth more than waiting and risking not getting it later.

What it means for businesses that depend on the Claude API

For companies in Latin America building products on top of Anthropic's API, this kind of deal is, at bottom, good news for continuity: it lowers the risk that Claude will hit severe capacity constraints or abrupt price hikes from compute scarcity in the near term. But it's also a signal that the structural cost of running these models remains enormous, which supports the expectation that frontier model pricing won't drop as fast as previous generations' did — it's worth continuing to design architectures that reserve the most expensive model for cases where it's genuinely justified.

Carlos Montiel
Enterprise AI Solutions Architect
Specialist in LLMs, Agents, and Orchestration
guatemalia.com/en/#contact · info@guatemalia.com

Need to implement AI at your company?

Carlos Montiel is an enterprise AI solutions architect. He implements LLMs, Agents, RAG, and orchestrators for companies in Guatemala and Latin America. Reach out for a consultation.

Contact Carlos Montiel

info@guatemalia.com