Shares of CoreWeave Inc. soared 8% in trading Wednesday after the specialized cloud provider announced a multi-year strategic partnership with artificial intelligence (AI) search pioneer Perplexity AI.
The deal greatly expands the neocloud operator’s footprint as it seeks to diversify its customer base beyond industry titans like OpenAI and Meta Platforms Inc.
Under the terms of the deal, Perplexity will deploy dedicated clusters powered by NVIDIA Corp.’s GB200 NVL72 Blackwell hardware to support its next-generation inference workloads. Inference, the process by which trained AI models generate real-time responses, is a rapidly growing sector of the market as AI applications move from the lab into production-scale consumer and enterprise tools.
The collaboration is structured as a two-way integration of technology.
Perplexity will leverage CoreWeave’s Kubernetes Service and high-performance computing clusters to power its Sonar and Search API ecosystem. Additionally, CoreWeave will implement Perplexity Enterprise Max across its own organization, allowing CoreWeave employees to utilize advanced AI models for deep multi-step research, data analysis, and unified internal document searching.
“This is a great win for CoreWeave and indicative of exactly the type of AI-native application that the company wants to build around as they expand their customer footprint beyond the frontier labs,” Dan O’Brien, president and chief operating officer of The Futurum Group, said on X. “This win should also alleviate some investor concerns about the useful life of its GPUs, as the inference counterweight to training workloads ensures strong lifecycle ROI on GPU investments.”
The announcement provides a much-needed boost for CoreWeave. Despite reporting a 168% year-over-year revenue jump to $5.1 billion in 2025, the company’s stock had recently faced downward pressure following a weaker-than-expected outlook and widening losses.
Investors remain wary of customer concentration, though CoreWeave’s $66.8 billion contracted revenue backlog suggests long-term demand for its GPU-as-a-service (GPUaaS) model.
“AI applications running in production require more than just access to raw infrastructure,” said Max Hjelm, CoreWeave’s senior vice president of revenue. “They require a cloud platform designed end-to-end for AI that simplifies compute operations.”
Dmitry Shevelenko, chief business officer at Perplexity, described CoreWeave as an “essential partner” for scaling the infrastructure behind its search products. In choosing a specialized provider like CoreWeave, Perplexity continues to execute a multi-cloud strategy while tapping into the latest Blackwell-based hardware from NVIDIA.
As the AI industry matures, the partnership highlights a shift in focus from mere model training to the delivery of reliable, high-speed inference. For CoreWeave, securing a fast-growing, AI-native client like Perplexity reinforces its position as the leading specialized alternative to traditional hyperscale cloud providers.
