Google will invest $10 billion in Anthropic PBC, with another $30 billion potentially to follow, strengthening a complex relationship between two companies that are at once partners and rivals in the race to build artificial intelligence.
The initial $10 billion cash injection values Anthropic at $350 billion, the same valuation the startup commanded in a funding round in February, a figure that notably excludes the capital raised in this latest tranche. According to the startup, the remaining $30 billion is contingent on Anthropic meeting specific performance milestones.
While the cash infusion is substantial, the broader agreement cements Anthropic’s status as a critical customer for Google’s cloud division. Alphabet Inc.’s Google Cloud has committed to providing 5 gigawatts of computing capacity over the next five years, with the potential for several more gigawatts to follow. This compute arrangement expands on a three-way deal announced earlier this month involving Anthropic, Google, and chip designer Broadcom Inc. (A single gigawatt is enough to power approximately 750,000 U.S. homes.).
For Anthropic, the deal addresses an urgent need for infrastructure to support exploding demand, particularly for its Claude Code agent, which automates software engineering tasks. The fundraising surge reflects the booming appetite for this product, which has driven the company to secure capital at a rapid clip.
“We need to build the infrastructure to keep pace with rapidly growing demand,” Anthropic CEO Dario Amodei said in a statement regarding the broader capacity expansion. The company is reportedly considering an initial public offering as soon as October.
Strategic Paradox
The deal deepens a longstanding yet paradoxical relationship. Anthropic relies heavily on Google’s Tensor Processing Units (TPUs), a scarce and precious resource considered the best alternative to Nvidia Corp.’s chips, to train its models. In return, Google secures a major anchor tenant for its cloud business, which is crucial as the company’s legacy search advertising business matures.
Yet the relationship is also intensely competitive. While Google is a key infrastructure supplier, its own Gemini* models are head-to-head rivals with Anthropic’s Claude. Current and former employees told Bloomberg that Google’s leadership has grown increasingly anxious about being outpaced in AI coding, a fast-growing segment where Anthropic has a commanding lead.
An Industry Template
The structure of the deal mirrors a similar agreement struck by the startup just days earlier with Amazon.com Inc. Amazon committed to invest $5 billion immediately, with an option to inject another $20 billion tied to performance, also at a $350 billion valuation. In return, Anthropic agreed to spend heavily on Amazon’s Trainium chips.
This circular dynamic, where tech giants invest in AI labs that use those investments to buy chips and cloud services from the same giants, has raised eyebrows among some financial analysts. However, for the moment, it has proven to be a powerful engine for growth.
Despite the windfall, Anthropic faces hurdles beyond its computational needs. The company has been deemed a “supply-chain risk” by the Pentagon, a designation it is currently fighting in court following a public dispute about how its technology could be used by the U.S. military.
