:

COREWEAVE LANDS MULTIBILLION-DOLLAR ANTHROPIC CONTRACT

AI DESK2 MIN READ
SUN, APR 12, 2026

■ AI-SUMMARIZED FROM 5 SOURCES BELOW

Cloud computing provider CoreWeave has signed a multibillion-dollar, multi-year deal with AI company Anthropic to develop and run its Claude AI models. The announcement comes alongside a separate $21 billion agreement with Meta.

CoreWeave announced Friday it secured a multibillion-dollar contract with Anthropic to provide cloud computing infrastructure for the AI company's Claude models. The multi-year partnership will see CoreWeave develop and operate the infrastructure needed to run Anthropic's AI systems. CEO Michael Intrator confirmed the deal's multibillion-dollar scope in an interview on Bloomberg Tech, though neither company disclosed exact financial terms. CoreWeave expects the first servers dedicated to Anthropic to come online later this year. The Anthropic deal caps a significant week for CoreWeave, which also announced a $21 billion agreement with Meta to supply computing power. The back-to-back announcements underscore the surging demand for specialized AI infrastructure as companies race to deploy and scale large language models. CoreWeave specializes in cloud computing services tailored for AI workloads, positioning itself as an alternative to traditional cloud providers. The company has focused on building infrastructure optimized for the intensive computational requirements of training and running AI models. Anthropic, the company behind the Claude family of AI assistants, has been expanding its infrastructure partnerships as it scales its operations. The deal with CoreWeave will provide dedicated computing resources for developing and deploying Claude models. Neither company provided details on the specific technical specifications of the infrastructure or the timeline beyond the initial servers launching this year. The partnership represents one of the larger known cloud computing deals in the AI sector.

■ SOURCES

Bloomberg TechThe InformationBloomberg TechTechmemeTechmeme

■ SUMMARY WRITTEN BY AI FROM THE LINKS ABOVE

■ MORE FROM THE AI DESK

A new benchmark reveals that even the best AI models struggle significantly with complicated visualizations. The RealChart2Code test shows leading proprietary models lose nearly 50% of their performance when handling complex charts built from real-world data.

JUST NOWAI Desk

Google has introduced A2UI 0.9, a framework-agnostic standard that enables AI agents to dynamically generate user interface elements. The standard works across web, mobile, and other platforms by leveraging existing app components.

1H AGOAI Desk

METR, a nonprofit AI research organization, has created benchmarks that are now widely used by researchers and investors to measure the pace of AI system development.

3H AGOAI Desk

Paris-based AI startup Mistral has pivoted from competing directly with OpenAI and Anthropic to positioning itself as an alternative to both Chinese and US AI labs. The company projects $80 million in monthly revenue by December.

4H AGOAI Desk

■ SUBSCRIBE TO THE DAILY BRIEF

ONE EMAIL, 5 STORIES, 06:00 UTC. UNSUBSCRIBE ANYTIME.