
Share
Anthropic and Amazon are ramping up their AI compute collaboration with a massive 5 gigawatt boost, set to power Anthropic’s Claude language model on custom Trainium chips, marking a monumental investment in AI infrastructure.
Anthropic, a leading AI research lab, has announced a significant expansion of its collaboration with Amazon to secure up to 5 gigawatts (GW) of new compute capacity. This agreement will support the training and deployment of Claude, Anthropic’s large language model, leveraging Amazon's custom Trainium2 and Trainium3 chips. The deal also includes substantial investments from both companies and a deeper integration of the Claude Platform within AWS.
The expanded partnership involves a commitment of over $100 billion over the next decade to AWS technologies, securing up to 5GW of new compute capacity for training and running Claude. This investment spans multiple generations of Amazon's custom silicon, including Graviton, Trainium2, Trainium3, and potentially future Trainium4 chips.
Anthropic will also use incremental capacity for Claude in Amazon Bedrock, expanding inference capabilities in Asia and Europe to better serve its growing international customer base. AWS remains Anthropic's primary training and cloud provider for mission-critical workloads.
The full Claude Platform will be directly integrated into AWS, offering a seamless experience for existing AWS users. This integration means:
This move ensures that organizations can leverage Claude while meeting their governance and compliance requirements. Claude remains unique in being available on all three of the world's largest cloud platforms: AWS (Bedrock), Google Cloud (Vertex AI), and Microsoft Azure (Foundry). The Claude Platform on AWS is set to launch soon, and interested users should reach out to their account teams for access.

Amazon has committed a substantial investment in Anthropic:
This builds on Amazon's previous investment of $8 billion, reflecting the growing importance and demand for Claude in various applications.
For AI practitioners, this collaboration offers several key benefits:
Tags
Original Sources
About the author
Kai built ML infrastructure at a Bay Area startup before developing an obsession with transformer architectures and inference optimisation that eventually pulled him out of product work entirely. A stint at a compute research lab sharpened his instinct for what actually matters in a model release versus what is marketing. He writes from the inside — from the perspective of someone who has debugged the systems he is describing at three in the morning. He is allergic to hype and instinctively drawn to the unglamorous plumbing questions that everyone else skips over.
More from The Engineer →This Week's Edition
21 April 2026
133 articles
Related Articles
Related Articles
More Stories