On April 24, 2026, Google made the largest single capital commitment to an AI company in recorded history. The company announced it would invest up to $40 billion in Anthropic, the San Francisco lab behind the Claude model family, with $10 billion arriving as an immediate cash transfer and the remaining $30 billion contingent on Anthropic clearing a set of undisclosed performance benchmarks. The deal lifts Anthropic's post-money valuation to $350 billion and locks in five gigawatts of dedicated computing capacity on Google Cloud infrastructure over the next five years.
How the Deal Is Structured
The two-tranche structure is unusual even by AI investment standards. Google commits to the first $10 billion now, with the second tranche of up to $30 billion available if Anthropic hits specific operational and financial targets, the details of which neither company has disclosed. What has been disclosed is the compute component: Google Cloud will provide five gigawatts of total capacity over five years, covering both the existing infrastructure already serving Claude's enterprise customer base and the additional headroom Anthropic needs as demand accelerates.
The deal also deepens an existing partnership. Claude has been available on Google Cloud Vertex AI since 2023, and the new agreement expands the compute and commercial relationship without altering Anthropic's multi-cloud stance. Claude continues to run on AWS Bedrock as well, reflecting Amazon's separate $25 billion commitment announced just four days before Google's announcement, on April 20. The two deals are structurally independent but commercially reinforcing: Anthropic now has more than $65 billion in committed cloud and compute backing from two of the three largest cloud providers.
Key Facts
- Upfront cash commitment$10 billion
- Total deal ceilingUp to $40 billion
- Post-money valuation$350 billion
- Google Cloud compute secured5 GW over 5 years
- Combined Google + Amazon commitmentOver $65 billion
- Enterprise customers spending $1M+/yr1,000+
The Numbers Behind the Commitment
The scale of the investment reflects what Google sees in Anthropic's current trajectory. Anthropic entered 2026 with a revenue run rate of approximately $9 billion at year-end 2025. By early April, that figure surpassed $30 billion on an annualized basis, a more than threefold jump in roughly three months. The growth is led by Claude Code, the company's agentic coding tool, which reached $1 billion in annualized revenue within six months of its public launch and has continued to compound. The number of enterprise customers spending more than $1 million per year with Anthropic crossed 1,000 in April, up from 500 when the company closed its Series G in February.
For Google, the investment serves more than one purpose. Google Cloud's business case depends in part on Anthropic's continued growth: every dollar of Claude API calls and every enterprise seat on Claude.ai represents workloads that can run on Google's infrastructure. Anthropic's revenue growth translates directly into revenue for Google Cloud's compute business. The investment is both a financial bet and a supply-chain commitment, and the structure blurs the line between customer relationship and equity stake.
"Our run-rate revenue has surpassed $30 billion, up from $9 billion at the end of 2025, as demand for Claude continues to accelerate. This partnership gives us the compute to keep pace." Anthropic, statement on the Google partnership, April 2026
Context Against the OpenAI Rivalry
The Google announcement arrived four days after Amazon confirmed its $25 billion infrastructure commitment, and it came against a backdrop of Microsoft's long-running OpenAI relationship, which now totals roughly $13 billion in committed capital. The arithmetic is direct: Anthropic's cloud and compute backers have committed more than five times what Microsoft has put into OpenAI. That comparison does not map cleanly onto company valuations or competitive dynamics, but it does signal where the largest technology companies are placing their largest bets.
For Anthropic, the deal resolves what had been the company's most significant near-term operating constraint. Running frontier AI models at scale is, above all else, an infrastructure problem. Demand for Claude grew faster in the first quarter of 2026 than Anthropic's internal planning had anticipated by a factor of eight, and the company's ability to serve that demand without degrading service quality depends on compute that has to be ordered years in advance. The Google and Amazon commitments secure that capacity through at least 2028.
What the Compute Timeline Means
Five gigawatts is a substantial number. For reference, Anthropic's current production infrastructure is estimated to require somewhere between 500 megawatts and 1 gigawatt to serve its existing traffic. The new Google Cloud capacity is sized for a company running several times larger than today, and it arrives in phases through 2027 and 2028 alongside the 3.5-gigawatt Broadcom TPU deal announced on the same April 6 day Anthropic crossed the $30 billion revenue run-rate threshold.
Anthropic said the infrastructure will be located primarily in the United States, aligning with a broader pattern of AI companies anchoring compute domestically as geopolitical pressure on supply chains increases. The specific Google Cloud regions were not disclosed, but the five-gigawatt figure implies construction of significant new capacity rather than reservation of existing rack space. Combined, the Google and Broadcom deals give Anthropic a combined pipeline exceeding 8 gigawatts of compute, to be deployed over roughly the same period that Anthropic's secondary market valuation crossed $1 trillion, signaling how rapidly market expectations have moved.
