Anthropic, the company behind the Claude assistant, received a further $4 billion commitment from Amazon and expanded its investment from Google, according to reporting by Reuters and CNBC. The company's Series E, disclosed later, valued it at roughly $35 billion.

The structure is the interesting part. Anthropic has sold cloud partnerships alongside equity: Amazon Web Services is the primary training partner for a defined class of models, and Google Cloud holds a similar role for others. FOUNDER. has not independently verified the contractual boundaries between the two.

Why two clouds

A single-cloud deal would have made Anthropic a captive supplier to one platform. Two deals let it negotiate compute pricing against itself while keeping its model roadmap independent of either company's product calendar.

The trade-off is governance. Anthropic's modified capped-profit structure, inherited from its founding team's time at OpenAI, means both investors accept return limits in exchange for proximity to a frontier model. That is a concession most large technology investors would not make for a smaller company.

What it changes

The cumulative capital puts Anthropic within reach of OpenAI's war chest, and its enterprise distribution through the two clouds gives it a route to large customers that bypasses the consumer product race. The question for competitors is whether a models-as-infrastructure business can stay independent once it depends on two hyperscalers for its existence.