Today, we published the results of our first fiscal year as a public company. It was a defining year for CoreWeave. But something else happened — a threshold crossed that sends a clear signal to the market.

In 2025, CoreWeave became the fastest cloud platform in history to surpass $5 billion in annual revenue. 

We are operating the largest AI-native cloud platform in the industry.

And that matters.

Over the past year, AI moved from experimental to essential. Enterprises and AI-native innovators alike began embedding AI into core operations. As workloads grew more complex and compute-intensive, a clear truth emerged: scaling AI requires a powerful cloud purpose-built for performance.

This is where CoreWeave stands apart. We power the intelligence driving the next wave of innovation. Our name is synonymous with the three ingredients customers need most right now—pace, performance, and partnership.  When infrastructure, software, and great teams work together, raw capacity turns into real-world outcomes.

Strip away all the complexity, and four fundamentals define where CoreWeave stands today. 

1. Relentless and Diversified Demand

Demand continues to broaden across hyperscaler, AI-native, and enterprise customers like Cognition, Cursor, Mercado Libre, Midjourney, and Runway, to name a few. In 2025, we grew revenue 168% year over year and expanded revenue backlog to $66.8 billion, providing multi-year visibility into expected future revenue.

The substantial majority of new 2026 capacity is already allocated under long-term commitments. Customers are no longer experimenting with AI, they’re operationalizing it. That shift from exploration to production is driving durable, diversified growth.

CoreWeave is supporting the next generation of AI pioneers. 

2. Cloud Services Beyond GPUs

CoreWeave is not simply deploying GPU capacity. Customers come to CoreWeave for our full-stack, AI-native cloud platform that integrates infrastructure, networking, orchestration, storage, and developer tooling into a unified system.

This platform increasingly defines the CoreWeave customer experience. In January, we announced  NVIDIA would test and validate our platform, including our software and reference architectures, working towards deepening interoperability with NVIDIA reference architecture to ultimately make these available to a broader customer set. 

Our differentiation lies not only in compute, but in how we operationalize it on CoreWeave Cloud.

We also introduced CoreWeave ARENA, a production-scale AI lab that enables teams to validate workload performance and economics before committing to scarce capacity. It underscores our conviction that performance is defined by orchestration and visibility, not compute alone.

Our approach has been recognized externally, reinforcing that we operate in a category of our own.

When infrastructure and software operate as one system, performance compounds. That integration defines CoreWeave’s advantage. 

3. Execution

Hyperscale is not defined by size alone. It is defined by the ability to execute at scale.

Bringing more than 260 megawatts online in a single quarter requires simultaneously orchestrating hardware, networking, storage, and purpose-built software across more than one hundred thousand GPUs and millions of interconnected system components. All in near-perfect unison. This is among the most operationally complex undertakings in the technology industry. It is also what CoreWeave does better than anyone. 

In 2025, that execution translated into 43 active data centers and more than 850 megawatts of active power. In the last three quarters alone, we added more capacity than in our entire prior history.

To put our current scale into perspective, according to third party estimates CoreWeave today is larger than the 15 largest neo-clouds across North America and Europe combined.

Even while operating at high utilization, we quickly resolved timing delays and returned to full deployment velocity. Execution at this scale requires coordination across power, hardware, and software.

I am immensely proud of how our teams always run toward the hard problems and deliver.

4. Disciplined Financial Model

Scaling AI requires more than technology. It requires capital discipline.

Our model is deliberately designed to invest against contracted demand. Backed by a $66.8 billion backlog, we have clear visibility into durable cash flows and attractive returns on capital.

Just as importantly, our close alignment with our customers gives us insight into how workloads are evolving and where AI is headed next. That vantage point informs where and how we deploy capital with precision rather than speculation.

We have built a capital framework that enables us to scale at the pace of AI adoption while maintaining balance sheet strength and progressing toward an investment-grade credit rating. As our weighted average cost of capital continues to decline, it further reinforces the durability of our model for sustained, disciplined growth aligned to the next phase of AI expansion.

—

This was a defining year for CoreWeave. We remain in the early stages of a generational infrastructure expansion. AI workloads are increasing in complexity, inference is scaling alongside training, and enterprise adoption continues to accelerate.

Our mandate is clear: continue to execute, raise the performance bar, and build the systems that power the next wave of innovation across the global economy.

When power and performance are no longer constraints, progress accelerates at scale. That is the work in front of us, the opportunity ahead is significant,  and we intend to lead it as The Essential Cloud for AI^TM.