Amazon has agreed to invest up to an additional $25 billion in the artificial intelligence startup Anthropic. This move builds on the $8 billion the company invested in recent years and will further expand their partnership to build vital AI infrastructure together.
According to an announcement from Amazon on Monday, Anthropic has committed to spending more than $100 billion on Amazon Web Services technologies over the next decade. This commitment includes using current and future generations of Trainium, which are custom AI chips designed by Amazon. Anthropic stated that it has secured up to five gigawatts of power capacity to train and deploy its Claude AI models.
Amazon Chief Executive Officer Andy Jassy noted in a statement that Anthropic committing to run its large language models on AWS Trainium over the next ten years reflects the progress made in their custom chip collaboration. He added that the company will continue providing customers with the necessary technology and infrastructure to build generative AI applications.
The investment from Amazon includes an immediate $5 billion tranche along with a further $20 billion tied to specific commercial milestones. This initial funding is based on Anthropic recently achieving a valuation of $380 billion. Anthropic expects to bring nearly one gigawatt of combined Trainium2 and Trainium3 capacity online before the end of this year.
As major cloud providers aggressively expand their capabilities, Amazon announced in February that its estimated capital expenditures for the year would reach approximately $200 billion. The company intends to allocate the majority of this budget to AI infrastructure.
This development notably follows Amazon agreeing just two months ago to invest up to $50 billion in OpenAI, the primary rival of Anthropic. Both AI firms are reportedly considering initial public offerings this year and are currently seeking to demonstrate their capabilities to investors. Executives at OpenAI recently criticized Anthropic for making a strategic error by failing to secure sufficient computing power.
Anthropic acknowledged on Monday that a rapid surge in demand for Claude from enterprises and developers along with increased consumer usage has placed inevitable pressure on its infrastructure. The startup noted that this pressure has impacted both reliability and performance. The new agreement with Amazon is expected to quickly scale its available computing capacity. Anthropic Chief Executive Officer Dario Amodei explained that users are increasingly finding Claude vital to their work, requiring the company to build infrastructure capable of keeping pace with this rapidly growing demand.
Founded in 2021 by a group of former OpenAI researchers and executives, Anthropic is known for its Claude models and has found initial success in the enterprise market. The startup recently surpassed an annualized revenue run rate of $30 billion. While Anthropic designated AWS as its primary cloud provider in 2023 and its main training partner in 2024, the company also maintains active partnership agreements with Microsoft and Google.