CoreWeave Signs Multi-Year GPU Deal With Anthropic, Its Second Major AI Infrastructure Agreement in 48 Hours

CoreWeave announced a multi-year agreement with Anthropic on April 10 to provide Nvidia GPU compute capacity across US data centers for production-scale Claude workloads. Financial terms were not disclosed. The deal arrives one day after CoreWeave announced a $21 billion expansion of its Meta partnership, making back-to-back announcements that position the company at the center of the AI infrastructure buildout.

Nine of Ten

Anthropic’s addition brings CoreWeave’s client roster to nine of the ten leading AI model providers, according to The Next Web. The company operates 32 data centers with more than 250,000 GPUs and 1.3 gigawatts of contracted power capacity. CoreWeave generated $5.13 billion in revenue in 2025 (a 168% year-on-year increase) and is guiding for more than $12 billion in 2026, backed by a contracted backlog exceeding $66 billion.

That growth has come with concentration risk. Microsoft accounted for approximately 67% of CoreWeave’s 2025 revenue, a dependence that investors flagged during the company’s March 2025 IPO. The Anthropic and Meta deals represent CoreWeave’s most visible effort to diversify beyond any single customer.

Anthropic’s Three-Lane Compute Strategy

Anthropic’s compute procurement now spans three distinct hardware architectures. Its primary training workloads run on Amazon Web Services Trainium chips via Project Rainier. Three days before the CoreWeave announcement, Anthropic signed a deal with Google and Broadcom for approximately 3.5 gigawatts of next-generation tensor processing unit capacity expected to come online in 2027, according to The Next Web. The CoreWeave deal fills a third lane: Nvidia GPU capacity for production inference.

The multi-vendor approach reflects the scale of demand. Anthropic’s annualized revenue run rate surpassed $30 billion in early April 2026, more than three times the $9 billion it recorded at the end of 2025, according to The Next Web. Claude Code adoption and enterprise deployments are the primary drivers of that acceleration, and inference compute is the bottleneck they create.

From Crypto Mining to AI Landlord

CoreWeave’s trajectory from Ethereum mining operation to critical AI infrastructure provider continues to accelerate. The company was founded in 2017 as Atlantic Crypto, buying Nvidia GPUs to mine cryptocurrency. When crypto margins compressed in 2019, it pivoted to GPU-on-demand cloud services. The AI training boom that began in 2023 turned that GPU stockpile into one of the most strategically valuable infrastructure positions in technology. CoreWeave went public on Nasdaq (CRWV) on March 28, 2025, at $40 per share, raising $1.5 billion at an approximately $23 billion valuation, according to The Next Web.

The Compute Bottleneck

The back-to-back Meta and Anthropic deals signal that GPU capacity is now the binding constraint on AI agent deployment at scale. Every major model provider needs more inference compute than any single cloud vendor can supply, and CoreWeave is positioning itself as the multi-tenant answer to that shortage. CRWV stock rose 3.93% in premarket trading on the announcement, according to Yahoo Finance.