Anthropic in talks with Google for a cloud compute deal worth high tens of billions

Summary

Anthropic is in talks with Google for a cloud compute deal valued in the high tens of billions, likely over $50 billion, according to Bloomberg reporting. The arrangement would provide Anthropic with additional computing power through Google's cloud services, though the discussions remain early and have not been finalized.

Google already owns roughly 41% of Anthropic and serves as one of its cloud providers. Expanding the relationship makes natural sense given Anthropic's business model. Anthropic generates revenue entirely through API-driven token sales to B2B customers, which means the company must continuously invest in compute to scale revenue and stay competitive. Unlike OpenAI, which could theoretically shift to profitability by monetizing user engagement through ads or commerce, Anthropic's survival depends on model innovation and inference capability.

The deal should be read as a scaling play rather than an acquisition precursor. Speculation that Google might acquire Anthropic outright misses the mark. Such a scenario would only emerge if Anthropic exhausted its ability to raise capital and faced pressure to cut spending. That does not appear imminent. Anthropic's business is described as doing fine and just growing.

Despite approaching three-quarters of OpenAI's revenue, Anthropic receives significantly less mainstream coverage. The gap stems mostly from the B2B-versus-B2C divide. OpenAI's ChatGPT reaches a billion weekly active users and makes for accessible headlines. Claude's user base consists largely of developers and enterprises. OpenAI also announces more product launches and partnerships. That visibility gap does not appear to bother Anthropic, which has maintained a quieter public profile while executing on core business metrics.