CoreWeave ( CRWV ) shares jumped nearly 6% in premarket trading on Wednesday after announcing a multi-year deal to support inference operations for Perplexity, an AI-powered search engine backed by Jeff Bezos and Nvidia.
As part of the agreement, CoreWeave will become a key cloud support partner for Perplexity AI. The company will run its next-generation tasks on dedicated NVIDIA GB200 NVL72 clusters managed by the cloud provider.
The platform will serve as the foundation for Perplexity’s Sonar and Search API products as they expand, the companies noted.
“AI applications running in production require more than just access to raw infrastructure – they require optimal performance and reliability, as well as a cloud platform designed for AI that simplifies computing operations,” said Max Hjelm, senior vice president of revenue at CoreWeave.
AI inference is the real-time execution phase of AI models, when trained models are used to make predictions or generate results based on new input data. This process can range from answering questions, making recommendations, classifying data to enabling real-time features such as search results, image recognition or language translation.
For the Perplexity product ecosystem, inference speed, latency stability, and scalability directly impact the user experience.
“We are proud to partner with Perplexity as they scale their workloads on CoreWeave’s AI cloud,” he said.
Dmitry Shevelenko, Perplexity’s Chief Commercial Officer, highlighted the provider’s technical capabilities and collaborative approach as key factors in the decision.
“We are impressed by the combination of CoreWeave’s technical capabilities and the first-rate thinking of our partners to help AI companies accelerate their growth and scale goals,” said Shevelenko, acknowledging CoreWeave’s role in enabling Perplexity to improve infrastructure efficiency and model quality to deliver powerful AI services in the search and automation sector.
The search firm has already started mining workloads using the Kubernetes cloud service provider. It also uses W&B models for training and fine-tuning as part of a broader cloud strategy.
Dedicated cloud GPU operators are becoming increasingly popular partners for AI companies facing ever-increasing computing demands. CoreWeave has posted leading results in MLPerf benchmarks and has platinum ratings in SemiAnalysis ClusterMAX evaluations for performance and reliability.
The deal will also see the cloud firm adopt Perplexity Enterprise Max internally, giving employees access to web search, research tools and advanced AI models through a single interface.
