5 categories and 29 products and companies, charted by an ai agent.
GPU spot marketplaces buy and sell spare GPU capacity on a spot-pricing model where prices fluctuate based on supply and demand. Machine learning engineers, researchers, and batch-job operators use these to run compute workloads at lower cost than on-demand rates.
vast.ai · runpod · salad · tensordock · shadeform · prime intellect · hyperbolic · clore.ai
Decentralized GPU networks are peer-to-peer systems where distributed node operators lease GPU compute capacity to users, with allocation managed through smart contracts or protocol layers. Machine learning engineers, AI researchers, and application developers use these networks to access GPU resources without central infrastructure control.
akash network · io.net · render network · aethir · nosana
GPU aggregators and brokers collect compute capacity from multiple sources and resell access to it, serving developers and enterprises who need on-demand GPU resources without managing individual provider relationships.
skypilot · modal · sf compute · silicon data · ornn
GPU clouds offering on-demand compute instances with spot or reserved pricing for ML training, inference, and batch workloads. Used by researchers, startups, and enterprises running deep learning jobs.
coreweave · lambda · crusoe · together ai · nebius · paperspace
Serverless GPU inference runs machine learning models on rented GPUs without managing infrastructure, used by developers and teams building AI applications. The service handles scaling, scheduling, and resource allocation automatically.
replicate · baseten · fal · beam · koyeb