Discover the cloud-first way to get the best of NVIDIA AI.
Try NVIDIA DGX Cloud Now
Explore NVIDIA NIM™ microservices on build.nvidia.com, a free API catalog for testing, prototyping, and developing generative AI apps with fully managed, accelerated endpoints and NVIDIA Blueprints — all accelerated by NVIDIA DGX Cloud.
Easily package and deploy inference pipelines or data preprocessing workflows in containers optimized for NVIDIA GPUs, without worrying about underlying infrastructure.
Fuel next-gen AI breakthroughs on NVIDIA DGX™ Cloud with NVIDIA GB200 featuring powerful NVIDIA Blackwell architecture, high-bandwidth NVIDIA NVLink™, and second-generation transformer engines that supercharge AI development and speed model pretraining.
NVIDIA DGX™ Cloud is a unified AI platform on leading clouds to optimize performance with software, services, and AI expertise for evolving workloads. Every layer of NVIDIA DGX Cloud is optimized and managed by NVIDIA, ensuring higher performance of NVIDIA in the cloud. A suite of fully managed platforms and services empowers every organization to bring AI workload from develop to deploy, in the era of agentic and physical AI. NVIDIA DGX Cloud includes: NVIDIA DGX Cloud Create, NVIDIA DGX Cloud Serverless Inference, and managed performance optimizations.
Use NVIDIA DGX Cloud Create on leading clouds, built with the latest NVIDIA technology and optimized for pre-training and fine-tuning, with flexible term lengths and seamless multi-cloud portability.
Use NVIDIA DGX Cloud Serverless Inference for production grade inference with auto-scaling, cost-efficient GPU utilization, and multi-cloud flexibility.
Try NVIDIA DGX Cloud Benchmarking Service, and follow evolving AI workload optimizations and workload-specific recipes to maximize your AI infrastructure performance.
Opt-in to NVIDIA NeMo Curator managed service to use NVIDIA DGX Cloud for video curation and post-training.
NVIDIA NIM™ provides prebuilt, optimized inference microservices that let you deploy the latest AI foundation models with security and stability on any NVIDIA-accelerated infrastructure — cloud, data center, and workstation. You can experience NVIDIA NIM on build.nvidia.com
Accelerated by NVIDIA DGX Cloud, build.nvidia.com is a model catalog with NIM microservices available through APIs and provides an environment for developers to try new software and models on NVIDIA DGX Cloud.