CoreWeave has completed its acquisition of Weights & Biases, the MLOps developer platform, closing a deal first announced in early March1. The transaction converts CoreWeave from a GPU cloud provider into a vertically integrated AI platform spanning compute, model training, experiment tracking, and application monitoring2.
CoreWeave said the combined entity will deliver what it calls the leading "AI Cloud Platform," purpose-built to develop, deploy, and iterate AI faster for AI labs and enterprises. The company framed the deal as a response to customer demand for end-to-end tooling suited to the inference era.
Weights & Biases is used as a system of record for training and fine-tuning AI models by organizations including OpenAI, Meta, NVIDIA, Snowflake, AstraZeneca, Toyota, Canva, Square, and Wayve. CoreWeave said the acquisition expands its customer base and provides "actionable insight into AI's highest-impact uses across the industry".
CoreWeave committed to maintaining interoperability of the Weights & Biases platform. Customers will retain the flexibility to choose any deployment option, any infrastructure provider, and any foundation model, framework, or protocol, according to the company. Weights & Biases users will be able to continue deploying workloads on-premise or with the infrastructure provider of their choice.
The unified platform is intended to deliver tighter AI workflow integration by combining infrastructure with experiment tracking, enabling faster and more efficient model development at scale. CoreWeave said the platform will provide real-time model performance monitoring and robust orchestration.
CoreWeave is also welcoming the Weights & Biases team as part of the transaction.
The deal adds a software layer to CoreWeave's infrastructure business, which recently integrated Run:ai's GPU orchestration platform across its cloud infrastructure for AI training and inference workloads[1]. ANALYSIS The Weights & Biases closure extends that vertical integration further up the stack, giving CoreWeave touchpoints from GPU scheduling through experiment management to production monitoring. The explicit interoperability commitment is notable: retaining multi-cloud and on-premise deployment options for Weights & Biases users reduces the risk that the acquisition drives customers toward CoreWeave-only infrastructure, a concern that typically accompanies platform acquisitions of developer tools.
CoreWeave had expected to close the acquisition in the first half of 2025, according to its prospectus dated March 27, 2025. The company filed that prospectus on March 31, 2025, pursuant to Rule 424(b) under the Securities Act of 1933.