ITMAP for Neocloud providers for AI

Neoclouds (AI Infrastructure Clouds) are specialized cloud providers purpose-built for artificial intelligence workloads, deriving the majority of their revenue from accelerated computing services — GPU or custom-silicon capacity for model training, fine-tuning, and inference. Unlike diversified hyperscalers, neoclouds focus exclusively on AI infrastructure, offering bare-metal clusters, managed Kubernetes/Slurm environments, and inference platforms with performance, pricing, and deployment flexibility optimized for AI builders. The market includes asset-heavy operators that own or lease data centers and accelerators, vertically integrated providers pairing proprietary silicon with cloud delivery, and asset-light platforms that orchestrate third-party capacity. Hyperscalers, colocation landlords, and pure capacity wholesalers are excluded.

Vendors and products in this market

• CoreWeave — CoreWeave Cloud Platform
• Crusoe — Crusoe Cloud
• Fireworks AI — Fireworks AI Inference & Fine-tuning Platform
• Fluidstack — Fluidstack GPU Cloud
• Lambda — Lambda Cloud (On-Demand & Reserved GPU Cloud)
• Modal — Modal Serverless GPU Platform
• Nebius — Nebius AI Cloud
• Nscale — Nscale Compute
• RunPod — RunPod GPU Cloud & Serverless
• Scaleway — Scaleway GPU Instances
• Tensorwave — Compute
• Together AI — Together GPU Clusters & Inference Platform
• Vast.ai
• Voltage Park — Voltage Park GPU Cloud
• Vultr — Vultr Cloud GPU

Explore the interactive vendor map and product comparisons. Registration and profile requirements apply to protected features, reports and intelligence.