CoreWeave reported first quarter 2026 revenue of $2.08 billion, up 112% year-over-year, as demand for GPU cloud infrastructure accelerated across AI labs, hyperscalers, and enterprise customers. The company said its revenue backlog reached $99.4 billion as of March 31, underscoring the scale of long-term commitments tied to AI model training and inference deployments. CoreWeave also surpassed 1 gigawatt of active power capacity and expanded contracted power capacity to more than 3.5 GW.
The quarter highlighted the rapid expansion of purpose-built AI infrastructure platforms as hyperscalers and AI-native companies race to secure compute capacity. CoreWeave disclosed multiple new agreements with Meta, including a new $21 billion commitment signed in March, alongside a multi-year agreement with Anthropic to support development and deployment of Claude AI models. The company also expanded relationships with Cohere, Mistral, Jane Street, Perplexity, Hudson River Trading, and World Labs. CoreWeave said it expects to scale toward more than 8 GW of power capacity by 2030 and separately announced an expanded partnership with NVIDIA to support deployment of more than 5 GW of AI factories.
CoreWeave continued to position itself as a vertically integrated AI cloud provider focused on inference and accelerated computing. During the quarter, the company launched Flexible Capacity Plans, including Flex Reservations and Spot offerings, introduced Dedicated Inference infrastructure services, and expanded the capabilities of its Weights & Biases platform for agentic AI and robotics development workflows. The company also secured an $8.5 billion delayed draw term loan facility structured as a non-recourse investment-grade financing vehicle and closed a $2 billion equity investment from NVIDIA. Despite the strong growth, CoreWeave posted a GAAP net loss of $740 million, driven in part by $536 million in quarterly interest expense associated with its aggressive infrastructure expansion.
- Q1 2026 revenue reached $2.078 billion, up from $982 million a year earlier
- Revenue backlog stood at $99.4 billion as of March 31, 2026
- Active power deployment exceeded 1 GW; contracted power surpassed 3.5 GW
- CoreWeave expanded total contracted power by more than 400 MW during the quarter
- Meta signed additional agreements including a new $21 billion commitment
- Anthropic signed a multi-year agreement supporting Claude model deployment
- NVIDIA invested $2 billion in CoreWeave equity
- CoreWeave secured an $8.5 billion DDTL 4.0 financing facility
- Adjusted EBITDA reached $1.16 billion with a 56% adjusted EBITDA margin
- Net loss widened to $740 million from $315 million a year earlier
“This was the strongest bookings quarter in CoreWeave’s history, with revenue backlog reaching nearly $100 billion,” said Michael Intrator, co-founder, chairman, and CEO of CoreWeave. “As the market moves from training to inference, that distinction matters more than ever. CoreWeave was built for exactly this.”
🌐 Analysis: CoreWeave’s latest quarter reinforces how AI infrastructure providers are evolving into strategic intermediaries between GPU vendors and AI model developers. The company’s nearly $100 billion backlog and long-duration commitments from Meta and Anthropic suggest that hyperscalers and frontier AI labs increasingly view external GPU cloud providers as essential capacity partners rather than temporary overflow infrastructure. CoreWeave’s emphasis on inference infrastructure, flexible reservation models, and vertically integrated software tooling also reflects the broader industry shift from massive pretraining clusters toward sustained production AI services.
🌐 Analysis: The company’s expansion beyond 1 GW of active power places it among the largest dedicated AI cloud infrastructure operators globally. The announcement also adds to a growing pattern of NVIDIA-backed infrastructure ecosystems that now include partnerships with firms such as CoreWeave, IREN, Nebius, and major colocation operators building AI factories optimized around NVIDIA accelerated computing platforms. Financing scale is becoming a key competitive differentiator, with CoreWeave relying heavily on structured debt markets and strategic equity backing to accelerate deployment timelines.
🌐 Analysis: CoreWeave’s first quarter capped a series of major strategic developments that significantly expanded its position in the AI infrastructure market. The most consequential was the company’s deepening relationship with Meta. During Q1, CoreWeave signed multiple new agreements with Meta, including a new $21 billion commitment disclosed in March. The deal highlighted Meta’s growing reliance on external GPU cloud infrastructure providers as it accelerates deployment of large-scale Llama training and inference environments. The agreement also reinforced a broader industry trend in which hyperscalers increasingly supplement internal AI infrastructure with specialized GPU cloud operators capable of deploying NVIDIA-based systems at exceptional speed and density.
🌐 Analysis: The Meta expansion followed a transformative period for CoreWeave that included its March 2025 Nasdaq listing, continued rapid buildout of GB200 NVL72 infrastructure, and expanded alignment with NVIDIA’s AI factory roadmap. During the quarter, CoreWeave surpassed 1 GW of active power capacity and announced plans with NVIDIA to support deployment of more than 5 GW of AI factories by 2030. The company also strengthened ties with Anthropic through a multi-year Claude infrastructure agreement while continuing to support leading AI-native firms such as Cohere, Mistral, and Perplexity. Collectively, these developments position CoreWeave as one of the largest independent AI cloud platforms operating between hyperscalers, frontier model developers, and the NVIDIA accelerated computing ecosystem.
🌐 We’re tracking the latest developments in AI infrastructure and hyperscale networking. Follow our ongoing coverage at: https://convergedigest.com/category/ai-infrastructure/
