CoreWeave is a specialized AI cloud infrastructure company. It sells access to GPU-rich cloud capacity and related platform software for large-scale AI training, production inference, agent development, data movement, and high-performance computing workloads. The company is focused on AI infrastructure rather than broad general-purpose cloud services.
The business model is built around committed cloud service contracts with customers that need large clusters of advanced GPUs, high-performance networking, and fast deployment. CoreWeave leases or licenses third-party data center capacity and invests heavily in servers, NVIDIA GPUs, networking, power infrastructure, software, and support teams. This creates a capital-intensive model with high upfront infrastructure spending and revenue recognized as contracted services are delivered.
In Q1 2026, CoreWeave reported revenue of $2.078 billion, up from $982 million a year earlier. Adjusted EBITDA was $1.157 billion, equal to a 56% adjusted EBITDA margin, showing strong infrastructure-level earnings before financing and depreciation effects. At the same time, the company reported a $144 million operating loss and a $740 million net loss, with $536 million of net interest expense. This reflects the scale of debt financing and capital investment required to build AI cloud capacity.
CoreWeave’s main revenue streams are
- Committed AI cloud capacity: The core revenue source is contracted access to GPU clusters and related infrastructure for customers running AI training, inference, and specialized compute workloads.
- Platform software and workflow tools: The company is expanding beyond raw compute through software used for AI development, monitoring, evaluation, and agentic workflows, including the W&B platform and ARIA research tooling.
- Enterprise and hyperscaler AI infrastructure: CoreWeave serves large AI labs, hyperscalers, enterprises, startups, and financial customers that need dedicated or large-scale AI compute capacity.
The company’s operating model is centered on AI factories and GPU cloud clusters across leased or licensed data center facilities in the United States, Europe, and the United Kingdom. Its active power footprint exceeded 1 GW in Q1 2026, and management is targeting more than 8 GW by 2030. CoreWeave also announced an expanded NVIDIA relationship intended to support the build-out of more than 5 GW of AI factories by 2030.
CoreWeave’s product and service categories include
- AI training infrastructure: Large GPU clusters for frontier model development and other compute-intensive training workloads.
- AI inference infrastructure: Cloud capacity for running production AI models at scale.
- High-performance computing: Specialized compute services for workloads that need dense accelerators and high-speed networking.
- AI developer workflow software: Tools tied to experimentation, model evaluation, observability, agent development, and research iteration.
- Data movement and platform services: Infrastructure services that support the movement, storage, and processing of large AI datasets.
CoreWeave’s competitive advantages are scale in specialized AI infrastructure, a strong NVIDIA-centered architecture, large committed contracts, and faster focus on AI-specific deployment than broad cloud platforms. Revenue backlog was $99.4 billion at March 31, 2026, giving the company a large base of contracted future revenue subject to service delivery and availability requirements. Management described Q1 2026 as its strongest bookings quarter, supported by major customer commitments.
Customer concentration is a defining feature of the model. In 2025, one customer represented 67% of revenue and another represented 15%. Significant disclosed customers include OpenAI, Microsoft, and Meta. This concentration supports rapid growth when large customers expand, but it also increases dependence on a small number of relationships.
CoreWeave competes directly with Amazon Web Services, Microsoft Azure, Google Cloud Platform, Oracle Cloud Infrastructure, and specialized AI cloud providers. Compared with AWS, CoreWeave is much narrower in scope. AWS offers a broad enterprise cloud platform across compute, storage, databases, applications, security, and analytics. CoreWeave is positioned as a specialist AI hyperscaler optimized for NVIDIA GPUs, high-performance clusters, and AI workload execution.
Its market position is strongest in the fast-growing AI infrastructure niche rather than the overall cloud market. The company has gained visibility through large AI customer agreements, a backlog approaching $100 billion, inclusion in the Nasdaq-100 Index in June 2026, and recognition as a Visionary in the 2026 Gartner Magic Quadrant for Cloud AI Infrastructure.
CoreWeave has limited disclosed direct exposure to China. The company reports revenue from the United States and all other countries, with 2025 U.S. revenue of $4.801 billion and all other countries revenue of $330 million. Its operations are concentrated in the United States, Europe, and the United Kingdom. China-related risk is mainly indirect through semiconductor supply chains, China-Taiwan tensions, export controls, tariffs, and access to advanced AI hardware.