Anthropic Commits $200B to Google Cloud and Chips Over Five Years

Anthropic has agreed to spend $200 billion with Google Cloud over the next five years, according to a report from The Information published Tuesday. The commitment, which begins ramping in 2027, would make Anthropic responsible for more than 40% of Google's recently disclosed cloud revenue backlog and stands as one of the largest cloud-and-chip contracts ever signed.

Alphabet shares jumped roughly 2% in extended trading following the report, as investors digested the scale of the long-term lock-in for Google's tensor processing units (TPUs) and cloud platform.

How We Got Here

The $200 billion commitment is the financial follow-through on a flurry of recent Anthropic-Google moves. In April, Anthropic signed a deal with Google and chip partner Broadcom for multiple gigawatts of TPU capacity expected to come online starting in 2027. Google has also agreed to provide up to 5 gigawatts of dedicated server capacity for Claude training and inference.

Alongside the infrastructure commitments, Alphabet is investing up to $40 billion directly into Anthropic, deepening a partnership that already saw Google contribute earlier rounds. The two companies now sit in an unusual position - simultaneously competitors in frontier model development through Gemini and Claude, and one of the most consequential supplier-customer pairs in the industry.

The deal lands at a moment when Anthropic is also fundraising heavily on the equity side. The company recently closed a round at a roughly $900 billion valuation, briefly making it the most valuable private startup in the world.

What's Inside the Deal

The $200 billion figure covers Google Cloud services and access to Google's custom AI accelerators - chiefly TPUs - rather than third-party GPUs from Nvidia. That detail matters. By committing this volume to TPUs, Anthropic is effectively betting that Google's silicon roadmap can match or beat what Nvidia can ship, both on price and on raw training throughput.

Reuters and The Information both report that the contract begins in 2027, which lines up with when the new gigawatts of TPU capacity are expected to come online. The agreement is structured as a multi-year purchase commitment, similar in shape to the long cloud contracts OpenAI has signed with Microsoft and Amazon Web Services.

Sherwood News notes that contracts involving Anthropic and OpenAI now account for more than half of the roughly $2 trillion in revenue backlogs at the major hyperscalers - AWS, Microsoft Azure, and Google Cloud. In other words, two AI labs are now the load-bearing customers for the entire enterprise cloud business.

Industry Impact

The most direct implication is for Google itself. The $200 billion commitment locks in years of high-margin AI revenue and validates the TPU strategy at exactly the moment when Wall Street has been asking whether custom silicon can really compete with Nvidia at frontier scale. Expect Alphabet's earnings calls for the rest of 2026 to lean heavily on this deal as a proof point.

It is also a signal to Nvidia and AMD. Anthropic was already a meaningful Nvidia customer through AWS, but the size of the Google-only commitment suggests the company is consolidating training workloads onto TPUs rather than fragmenting across accelerators. If a top-two frontier lab can run primarily on TPUs, other developers may revisit the assumption that Nvidia is the only viable path.

For enterprise buyers, the deal reinforces a now-familiar dynamic: AI capacity is being pre-allocated years in advance to a small number of frontier labs. Mid-sized customers shopping for premium GPU capacity in 2027 and 2028 will increasingly find it spoken for.

What People Are Saying

CNBC's "Fast Money" desk discussed the trade Tuesday, with traders flagging Alphabet as the clearest beneficiary and noting that the news partially explains why Google has held its ground in the AI infrastructure narrative even as questions persist about Gemini's consumer momentum.

Engadget framed the deal as a strategic hedge for Anthropic. "Going all-in on Google's chips means Anthropic doesn't have to fight the rest of the industry for Nvidia allocation," the site noted, pointing out that the company's compute needs have grown faster than the market can supply general-purpose GPUs.

Some analysts have raised eyebrows at the circular nature of the relationship - Google invests in Anthropic, Anthropic spends that capital (and more) on Google services, Google books the revenue. It is a pattern that increasingly defines the AI economy, and one regulators are starting to watch.

What's Next

Watch for two things in the coming weeks. First, official confirmation. Neither Anthropic nor Google has commented publicly on the specifics, and the deal's structure may shift before any formal announcement. Second, Nvidia's response. Jensen Huang has been vocal about the staying power of GPUs for training, and a $200 billion TPU commitment from a frontier lab is the kind of news that typically prompts a counter-narrative.

For developers using Claude, the practical impact will arrive gradually. More dedicated capacity should mean fewer rate limits and more predictable inference performance for enterprise customers, especially as the 2027 gigawatts come online.

Bottom Line

This is the largest concrete expression yet of how Anthropic plans to fund its compute future - and it leans hard into Google rather than diversifying. For Alphabet, it is validation of the TPU bet at unprecedented scale. For the rest of the industry, it is a reminder that the AI build-out is being financed in trillion-dollar commitments, and the seats at that table are filling up fast.