Technology

A Cloud Company Called Nscale Is Buying Anyscale for $1.65 Billion to Build an All-in-One AI Platform

Martin HollowayPublished 19h ago4 min readBased on 5 sources
Reading level
A Cloud Company Called Nscale Is Buying Anyscale for $1.65 Billion to Build an All-in-One AI Platform

Nscale is acquiring Anyscale for $1.65 billion. Nscale provides the hardware that powers artificial intelligence, and Anyscale makes the software that manages how that hardware gets used. Together, they cover both pieces of what companies need to run AI systems. The deal was publicly announced on Thursday, July 30, 2026 (Reuters). Bloomberg and Reuters each cited a person familiar with the matter for the purchase price, as reported by TechCrunch.

Nscale published its own press release on its domain the same day, titled "Nscale Acquires Anyscale, Enhancing its Full Stack AI Cloud Platform" (Nscale). Anyscale confirmed the agreement in a blog post on its own site. The companies said the acquisition would help clients better manage and run AI systems efficiently, per Reuters.

Think of it like buying a factory and then buying the management software that schedules the machines. Nscale already owns the factory — the powerful computers, called GPUs, that do the heavy lifting for AI. Anyscale makes the software that decides which tasks go to which machines and keeps everything running smoothly.

Anyscale's software is built around something called Project Ray, a free, open-source tool that helps developers split big computing jobs across many machines at once. The company was founded by the same team that created Project Ray, giving it deep expertise in that problem. After the launch of GPT-3 in 2022, Anyscale shifted from general-purpose computing to focus on AI tasks: training models, serving them to users, curating data, generating outputs, and reinforcement learning.

That shift appears to be paying off. Anyscale's revenue increased 70 percent in its most recent quarter compared to the one before it. Anyscale was valued at $1.38 billion in a 2022 funding round, making the $1.65 billion acquisition price a premium over that last private valuation, though the intervening years and revenue growth complicate direct comparison.

Nscale brings the other half. The company raised $2 billion in March 2026 at a $14.6 billion valuation. Its investors include Nvidia, Nokia, Blue Owl, Dell, and Norwegian industrial group Aker. That capital base gives Nscale the financial strength for a deal of this size, and the investor list signals alignment across the hardware supply chain, from chips to systems to infrastructure.

Under the acquisition terms, Anyscale will continue to operate under its own branding and serve existing customers. All of Anyscale's approximately 200 employees will join Nscale. The preservation of the Anyscale brand and customer relationships matters here: Project Ray has a substantial open-source community of users who chose the tool precisely because it is not locked to a single cloud provider. How Nscale navigates that open-source trust while integrating Anyscale's commercial layer into its own infrastructure will be one of the deal's biggest challenges.

For Nscale, the acquisition means moving beyond selling raw computing capacity to offering the software that makes that capacity usable for AI work. For Anyscale, it means attaching its software to a dedicated, well-funded hardware provider rather than running on other companies' computers.

Cloud providers have long understood that owning both the hardware and the software on top of it increases customer loyalty, profit margins, and differentiation. AWS built SageMaker; Google Cloud built Vertex AI; Azure built its AI Foundry. Nscale is a younger, AI-focused company attempting the same strategy from a smaller starting point, and it is buying rather than building the software layer.

The broader question is whether the AI infrastructure market will support another full-stack player alongside the giant cloud providers — Amazon, Google, and Microsoft — or whether the field consolidates around a few providers big enough to compete on cost. Nscale's March 2026 funding and this acquisition are bets that a company purpose-built for AI can win customers by being specialized rather than general-purpose. Anyscale's 70 percent revenue growth suggests real demand for the software layer. Whether that demand flows to an independent provider or gets absorbed into the giants' own offerings is the open question this deal poses.

For Anyscale's approximately 200 employees and existing customers, the immediate transition appears minimal: same brand, same product, same relationships, now backed by a $14.6 billion-valued parent. The longer-term test will be whether Nscale can resist the temptation to tighten Anyscale's platform around its own infrastructure at the expense of the open, flexible approach that made Project Ray and Anyscale attractive in the first place.