Amazon is expanding its Amazon Anthropic alliance with a new investment and a much larger infrastructure commitment, underscoring how central cloud capacity has become in the AI race. Reuters reported that Amazon will invest $5 billion in Anthropic immediately and up to another $20 billion later, while Anthropic has agreed to spend more than $100 billion over the next 10 years on Amazon Web Services technologies. Amazon and Anthropic said the new agreement also gives Anthropic access to up to 5 gigawatts of compute capacity to train and run its Claude models.
The deal pushes the relationship well beyond a simple funding round. It ties Anthropic more deeply to Amazon’s cloud and chip ecosystem, especially AWS Trainium and Graviton, at a time when frontier AI companies are competing for scarce computing power. It also gives Amazon another way to strengthen AWS against Microsoft and Google in the market for AI infrastructure and enterprise model access.
What the Amazon Anthropic Alliance Includes
Amazon said Anthropic will commit more than $100 billion over the next decade to AWS technologies, spanning current and future generations of Trainium chips from Trainium2 through Trainium4 as well as tens of millions of Graviton CPU cores. Anthropic said the agreement secures up to 5 gigawatts of capacity for training and deploying Claude, with additional Trainium2 capacity arriving in the first half of 2026 and nearly 1 gigawatt of Trainium2 and Trainium3 capacity expected to be online by the end of the year.
Amazon’s direct financial commitment is also significant. Amazon said it is investing $5 billion now and could invest up to another $20 billion if certain commercial milestones are met. That comes on top of the roughly $8 billion Amazon had already invested in Anthropic before this latest deal.
Another key part of the agreement is product integration. Amazon said AWS customers will be able to access the full Anthropic-native Claude console from within AWS, making it possible to use Claude Platform through an existing AWS account without separate credentials, contracts or billing. For enterprise customers, that lowers friction and makes AWS a more complete home for Anthropic-based AI workloads.
Why the Partnership Matters for Both Companies
For Anthropic, the agreement is largely about guaranteed infrastructure. Training and serving advanced AI models now requires enormous computing resources, and access to that capacity has become one of the biggest constraints in the industry. Anthropic said demand for Claude has accelerated sharply in 2026 and that its annualized revenue run rate has risen above $30 billion, up from about $9 billion at the end of 2025.
That growth has created pressure on reliability and performance. Anthropic said the new agreement will expand available capacity quickly, with meaningful new compute arriving in the next few months. That gives the company a clearer path to scaling Claude across enterprise and consumer products without depending on short-term capacity deals.
For Amazon, the value of the relationship goes beyond its equity stake. The expanded alliance gives AWS a marquee AI customer that is committing enormous long-term spending to Amazon’s cloud and in-house silicon. Amazon CEO Andy Jassy said Anthropic’s long-term commitment to Trainium reflects the progress the two companies have made together on custom silicon. That matters because AWS has been trying to prove that Trainium and Graviton can compete as lower-cost, high-performance options for large-scale AI training and inference.
Why This Matters for the Wider AI Market
The agreement also highlights a bigger shift in the AI economy. Capital alone is no longer enough. The companies with the best access to chips, power, data center capacity and cloud distribution are gaining an edge. Reuters has described the Amazon-Anthropic arrangement as part of a broader wave of multibillion-dollar AI, cloud and chip deals that are reshaping the sector.
The deal does not make Anthropic exclusive to Amazon. Anthropic said Claude remains available through AWS, Google Cloud and Microsoft Azure. Even so, Anthropic also said AWS remains its primary training and cloud provider for mission-critical workloads, which gives Amazon a meaningful edge in one of the most valuable parts of the AI market.
There is also a broader structural implication. Deals like this help fast-growing AI firms secure the infrastructure they need, but they also deepen the dependence of model developers on a small number of hyperscale cloud providers. That does not eliminate competition, especially since Anthropic still works across multiple platforms, but it does show how tightly the future of AI development is becoming linked to the largest cloud and chip ecosystems.
What Comes Next
In practical terms, the next phase of the alliance will be measured by rollout and usage. Anthropic said meaningful new capacity will come online quickly, with larger Trainium3 deployments expected later this year. Amazon is expected to use the relationship to attract more enterprise customers to AWS, Bedrock and its custom silicon offerings.
The alliance also shows that Amazon intends to stay aggressive in AI without relying on a single in-house model strategy. By deepening its relationship with Anthropic, Amazon is betting that control over infrastructure, cloud distribution and enterprise integration can be just as important as owning the model itself. Anthropic, for its part, is making a long-term bet that Amazon can provide the scale and reliability it needs for Claude’s next stage of growth.
