Anthropic and Amazon Extend Collaboration to Scale Claude Across AWS and Beyond

Page 3

Anthropic has announced a strengthened agreement with Amazon to expand their long-running partnership and scale Claude across AWS, aiming to secure up to 5 gigawatts of capacity for training and deploying Claude. The deal also brings new Trainium2 capacity online in the first half of this year, with nearly 1 GW of Trainium2 and Trainium3 capacity expected by the end of 2026. Since 2023, more than 100,000 customers have used Claude on Amazon Bedrock, and the collaboration has already powered large compute projects like Project Rainier—the team-wide compute cluster—and currently leverages over one million Trainium2 chips to train and serve Claude.

Key elements of the expanded agreement

Infrastructure at scale: Amazon plans to invest more than $100 billion over the next decade in AWS technologies, securing up to 5 GW of new capacity to train and run Claude. The commitment covers multiple generations of Amazon’s custom silicon, from Graviton to Trainium2 through Trainium4, with an option to adopt future chips as they become available.
Capacity ramp and product scope: A substantial portion of Trainium2 capacity is expected to come online in Q2, with scaled Trainium3 capacity anticipated later this year. Anthropic will also use incremental capacity to run Claude on Bedrock, while expanding Claude’s inference capacity in Asia and Europe to better serve a growing international customer base.
Cloud strategy: Anthropic intends to continue treating AWS as its primary cloud and training provider for mission-critical workloads. The companies emphasize joint progress on custom silicon and the infrastructure required to support rapid AI innovation.

“Amazon’s custom AI silicon offers strong performance at a lower cost for customers, driving real demand,” said Andy Jassy, CEO of Amazon. “Anthropic’s decision to run its large language models on AWS Trainium for the next decade reflects how far we’ve come together—providing the technology and scale customers need to build with generative AI.”

Claude Platform on AWS

The full Claude Platform will be available directly within AWS, accessible from the same account with the same controls and billing, and without the need for new credentials or contracts. This integration brings Claude into a familiar governance and compliance framework for organizations, while maintaining interoperability across platforms. Claude remains the only frontier AI model available on all three major cloud platforms: AWS Bedrock, Google Cloud Vertex AI, and Microsoft Azure Foundry. Claude Platform on AWS is coming soon; interested teams are encouraged to contact their AWS account representative to request access.

Continued investment and growth

Amazon is committing $5 billion to Anthropic now, with the potential for as much as $20 billion more in the future, complementing the company’s prior investment totaling $8 billion. Dario Amodei, CEO and co-founder of Anthropic, notes that Claude’s growing importance for customers drives the need for robust infrastructure. The collaboration with Amazon will enable ongoing AI research and broader access to Claude for the more than 100,000 organizations already building on AWS.

Meeting record demand and expanding capacity

Anthropic reports rising demand for Claude among enterprises, developers, and consumers across free, Pro, and Max tiers. The company’s run-rate revenue has surpassed $30 billion, up from roughly $9 billion at the end of 2025. To keep Claude at the AI frontier and maintain reliability amid rapid growth, the alliance will accelerate capacity deployment over the next three months and push toward nearly 1 GW of additional capacity by year’s end. A diversified hardware strategy—spreading workloads across multiple chip families—will help meet peak demand and ensure consistent performance.

To learn more about Anthropic on AWS, visit: https://aws.amazon.com/bedrock/anthropic/

Updated April 21 to note that Claude Platform on AWS is coming soon.

Anthropic and Amazon Extend Collaboration to Scale Claude Across AWS and Beyond

Related links