Lightning AI on 19 Aug 2025 unveiled a multicloud GPU Marketplace that lets developers compare and rent graphics processors from both major hyperscalers—such as AWS and Google Cloud—and specialized “neocloud” providers including Lambda and Voltage Park. The service is embedded in the Nvidia-backed startup’s end-to-end AI development platform, giving users a single interface to run training, fine-tuning and inference workloads without rewriting code or rebuilding their infrastructure. Chief Executive Officer William Falcon said the launch is meant to give AI teams “control, flexibility and speed,” while Chief Product Officer Priya Shivakumar described the goal as becoming the “iOS of AI development platforms.” Lightning AI claims the marketplace can cut infrastructure bills by as much as 70% through transparent, side-by-side pricing for spot, on-demand and reserved capacity. Developers can start free; the lowest paid tier costs $140 per user per month, with enterprise plans negotiated separately. Demand for GPUs has surged amid a global shortage driven by large-scale generative-AI projects, pushing companies to seek new ways to secure affordable compute. By aggregating supply across multiple clouds and eliminating vendor lock-in, Lightning AI aims to position itself as a neutral broker in a market long dominated by the biggest cloud providers.