Amazon’s $25B Anthropic Investment: AWS and 5 GW of Compute

Five billion now, up to twenty more tied to milestones

Amazon agreed to invest up to $25 billion more in Anthropic: $5 billion immediately, priced at Anthropic's $380 billion valuation from its February 2026 funding round, with up to $20 billion more tied to unspecified commercial milestones. Combined with the $8 billion Amazon had already put into Anthropic, its total commitment reaches up to $33 billion. [1][2]

In exchange, Anthropic committed to spending more than $100 billion on Amazon Web Services technologies over the next decade, including current and future generations of Amazon's custom Trainium chips, and secured up to 5 gigawatts of capacity for training and running Claude. Anthropic said it would bring nearly 1 gigawatt of combined Trainium2 and Trainium3 capacity online by the end of 2026. [1][2][3]

Claude outgrew the infrastructure it had

Anthropic said enterprise and developer demand, plus a "sharp rise" in consumer usage, had produced what it called "inevitable strain" on its infrastructure, hurting Claude's reliability and performance. Its annualized revenue had passed $30 billion, up from roughly $9 billion at the end of 2025. [1][3]

Amazon CEO Andy Jassy said Anthropic's commitment to run its large language models on AWS Trainium for the next decade reflected the progress the two companies had made on custom silicon. Anthropic CEO Dario Amodei said the company needed to build infrastructure to keep pace with growing demand while continuing to serve the more than 100,000 customers already building on AWS. [1][2]

One of three big cloud provider expansions inside three months

The deal landed two months after Amazon agreed to invest up to $50 billion in OpenAI, Anthropic's chief rival, as part of a $110 billion round that valued OpenAI at $730 billion, and in the same month Anthropic expanded its own arrangements with Google and Broadcom for what it called "multiple gigawatts" of additional capacity. Anthropic had also agreed in November 2025 to a Microsoft investment of up to $5 billion alongside a $30 billion Azure purchase commitment. [1]

Eight days later, OpenAI's own models became available to run on AWS as its exclusivity arrangement with Microsoft ended, putting Amazon in business with both labs on the same cloud within the same quarter it funded them. [1]

Atlas interpretation: Read together, the run of deals shows the major clouds funding both of the best capitalized AI labs at once rather than picking a side, and buying compute commitments and revenue from each in the process. TechCrunch's coverage describes Amazon's investment as racing to keep pace with OpenAI's own infrastructure buildout, not as evidence about either company's next funding round. [4]