A recent study conducted by Qualitate, based on 13 in-depth interviews with senior decision-makers, provides insights into CoreWeave's market position. The research panel included representatives from large enterprises across sectors like healthcare, technology, and finance, located in the US, Canada, and Europe.
Of the 13 participants, seven currently use CoreWeave, with annual spending ranging from below $100,000 to over $5 million. All seven current users plan to expand their workloads or spending. The remaining participants include those still evaluating the platform, one who completed a pilot but did not launch, and two who opted for other providers.
Ten respondents cited friction with hyperscaler GPU capacity or product access as a primary factor in their decision to look elsewhere. Customers highlighted CoreWeave's faster deployment times, superior inference latency, and stability as key operational advantages that encourage long-term retention.
The data indicates that AI inference workloads are growing alongside training, with some customers shifting their training-to-inference mix toward a 50-50 split. While CoreWeave captures significant AI workloads, all current users maintain broader relationships with hyperscalers, often using them for application services while hosting models on CoreWeave.
CoreWeave is expanding its portfolio through acquisitions like Weights and Biases and the development of its Omni AI stack for private data centers. The company differentiates itself from other GPU clouds through enterprise support, cluster availability at scale, and operational flexibility.
