Nscale Acquires Anyscale for $1.65 Billion to Own AI Stack

British AI neocloud Nscale is acquiring software startup Anyscale for $1.65 billion, a move that extends the company's grip across the AI infrastructure stack — from energy and data centers all the way up to workload management software.

What Happened

Nscale, which raised $2 billion in a Series C round this March at a $14.6 billion valuation, confirmed the deal on Wednesday. Anyscale, the company behind the open-source Project Ray distributed computing framework, helps organizations scale AI workloads — including large language model training, inferencing, data curation, and reinforcement learning — across data centers and servers.

The acquisition gives Nscale a software layer it previously lacked. The company has been building vertically, establishing business lines in energy supply, data center operations, and orchestration software. With Anyscale, it adds workload management and scaling directly into its offering.

Nscale's investors — including Nvidia, Nokia, Blue Owl, and Dell — are backing a company that increasingly resembles a full-stack alternative to hyperscalers, but purpose-built for AI workloads. The neocloud has been aggressively deploying capital, securing partnerships with Microsoft, British Telecom, and Nordcraft for compute and data center capacity.

Why It Matters

Anyscale was valued at $1.38 billion in its 2022 Series C round and reported 70% quarter-over-quarter revenue growth in its most recent quarter. The startup's approximately 200 employees will join Nscale, and Anyscale will continue operating under its own brand.

"Together, Anyscale and Nscale can co-design the software layer and infrastructure beneath it, something that neither company could do as effectively by optimizing its layer alone," Anyscale said in a statement.

What Builders Should Watch

For AI builders and technical teams, this acquisition signals that the infrastructure layer is maturing fast. Nscale's vertical integration — spanning power, physical data centers, orchestration, and now workload management — points toward a future where AI compute is bought as a fully managed service rather than assembled from separate vendors.

The combined entity will likely compete more aggressively on pricing and developer experience, which could lower the barrier for teams training or serving large models. Anyscale's roots in Project Ray also mean the open-source community should watch how Nscale handles the project's governance and roadmap going forward.

The deal reflects a broader consolidation trend in AI infrastructure, where cloud providers and neoclouds are racing to offer end-to-end solutions — a race that shows no signs of slowing.