Author: Claude, Shenchao TechFlow
DeepChao Summary: On Monday, Amazon announced an additional investment of up to $25 billion in Anthropic (with $5 billion immediately available) and secured a commitment from Anthropic to spend over $100 billion on AWS over the next decade.
This is the second time in two months that Amazon has issued a billion-dollar check to a leading AI lab—following its recent $50 billion investment in OpenAI.
Anthropic's annual revenue has surpassed $30 billion, but compute constraints are hindering user experience; the core objective of this transaction is to address the capacity crisis.
Amazon is making increasingly large bets on two major AI labs.
According to multiple media outlets including CNBC and Bloomberg on April 20, Amazon announced an additional investment of up to $25 billion in Anthropic, with $5 billion to be paid immediately and the remaining $20 billion tied to specific business milestones. This investment is based on Anthropic’s $380 billion valuation from its G-round financing in February, and, when combined with Amazon’s prior cumulative investment of $8 billion, Amazon’s total committed investment in Anthropic now reaches a maximum of $33 billion.
Two months ago, Amazon invested $50 billion in OpenAI, Anthropic’s primary competitor, and entered into a cloud services agreement of similar scale. Amazon CEO Andy Jassy stated in a statement that Anthropic’s commitment to running its large language models on AWS Trainium for up to ten years “reflects our progress in custom chips.”
After the announcement, Amazon's stock rose approximately 2.5% in after-hours trading.
$100 billion cloud commitment for 5 gigawatts of computing power, responding to OpenAI's "insufficient computing power" allegations
At its core, this transaction is not just an equity investment, but a deeply integrated infrastructure agreement.
Anthropic has committed to investing over $100 billion in AWS technologies over the next decade, encompassing Amazon’s custom AI chips Trainium (from Trainium2 through Trainium4 and future generations) and tens of millions of Graviton CPU cores. In exchange, Anthropic will receive up to 5 gigawatts of computing capacity for training and deploying Claude models. According to Anthropic’s blog, the company has already used over one million Trainium2 chips to train and serve Claude, and plans to bring nearly 1 gigawatt of Trainium2 and Trainium3 capacity online by the end of 2026.
This expansion in computing power directly responds to recent public criticisms from OpenAI. Last week, in an internal memo, OpenAI’s Chief Revenue Officer Denise Dresser claimed that Anthropic had made a “strategic misstep by failing to secure sufficient compute,” and predicted that OpenAI would have 30 gigawatts of compute by 2030, while Anthropic would have only 7 to 8 gigawatts by the end of 2027. In its announcement that day, Anthropic acknowledged that demand for Claude from enterprises and developers has accelerated rapidly, with consumer usage also experiencing a “sharp increase,” placing “unavoidable pressure” on its infrastructure and impacting reliability and performance during peak hours.
Anthropic CEO Dario Amodei stated in a statement: “Users have told us that Claude is becoming increasingly essential to how they work, and we need to build infrastructure to keep up with rapidly growing demand.”
Amazon issued billion-dollar checks to two AI labs within two months.
Amazon's investment strategy is now clear: betting on both leading players in the AI赛道.
In February, Amazon announced an investment of up to $50 billion in OpenAI, accompanied by a $100 billion commitment to AWS cloud services. Today’s deal structure with Anthropic is nearly identical—$25 billion in investment plus over $10 billion in locked-in cloud spending. According to GeekWire, Amazon is executing the same playbook with both labs.
Two major AI companies are also competing to demonstrate their strength to investors. According to CNBC, both Anthropic and OpenAI are preparing for an IPO that could occur as early as this year. OpenAI’s most recent funding round valued the company at over $850 billion, while Anthropic was valued at $380 billion. Anthropic claims its annualized revenue has surpassed $30 billion (approximately $9 billion by the end of 2025), but OpenAI’s internal memo alleges that this figure is inflated by about $8 billion because Anthropic includes the full amount of cloud partnership revenues from Amazon and Google rather than reporting net revenue.
Microsoft is also betting on both sides—having already invested over $13 billion in OpenAI, it further invested up to $5 billion in Anthropic in November 2025, with Anthropic committing to purchase $30 billion in Azure computing power.
Claude platform joins AWS in a battle for over 100,000 customers
Beyond investment, the integration between both parties on the product level is also deepening.
According to the announcement, Anthropic’s native Claude platform will be directly integrated into AWS, allowing users to access the full Claude console through their existing AWS accounts, permission controls, and billing systems—without requiring additional registration or new contracts. This represents an advancement over the previous method of offering Claude via the Amazon Bedrock marketplace. Amazon disclosed that over 100,000 organizations are currently running Claude models on Amazon Bedrock.
Anthropic also emphasized in its blog that Claude is the only cutting-edge AI model available simultaneously across the three major cloud platforms: AWS Bedrock, Google Cloud Vertex AI, and Microsoft Azure Foundry. This multi-platform strategy allows enterprise customers to flexibly choose their deployment options and represents one of Anthropic’s key differentiators in competition with OpenAI.
On the client side, Lyft reduced its average customer service resolution time by 87% by using Claude, powered by Amazon Bedrock, to drive its customer service AI assistant. Pfizer uses Claude to help scientists perform voice searches within drug development documents, saving approximately 16,000 hours of search time annually.
AI Infrastructure Race: Amazon's Capital Expenditure Is Expected to Reach $200 Billion This Year
The broader context of this transaction is the AI infrastructure arms race among cloud computing giants.
In February, Amazon stated that it expects capital expenditures to reach approximately $200 billion by 2026, with the vast majority allocated to AI infrastructure. Previously, the joint Project Rainier—a massive computing cluster with nearly 500,000 Trainium2 chips—was among the world’s largest AI computing clusters, and Anthropic is using it to train and deploy current and future versions of Claude.
Earlier this month, Anthropic also expanded its collaboration with Google and Broadcom to secure gigawatt-scale computing power, expected to come online starting in 2027. Combined with this new 5-gigawatt agreement with Amazon, Anthropic is simultaneously expanding its computing capacity across multiple fronts.
Amazon’s custom chip business is also accelerating. Jassy recently revealed that the business’s annualized revenue has surpassed $20 billion, doubling from the $10 billion reported earlier this year—describing it in his own words as “exploding.”