NVIDIA has introduced DGX Cloud Lepton, a new AI platform designed to connect developers with tens of thousands of GPUs through a global marketplace. This platform will support a wide range of agentic and physical AI applications by providing scalable access to GPU compute power from cloud providers around the world.

To meet the rapidly growing demand for AI, NVIDIA has partnered with several cloud providers known as NVIDIA Cloud Partners (NCPs) including CoreWeave, Crusoe, Firmus, Foxconn, GMI Cloud, Lambda, Nebius, Nscale, Softbank Corp., and Yotta Data Services. These partners will offer NVIDIA’s latest Blackwell GPUs and other architectures through the DGX Cloud Lepton marketplace.

What DGX Cloud Lepton Offers?

DGX Cloud Lepton provides developers access to GPU capacity on-demand or for long-term use, based on their needs and regional preferences. This flexibility is especially valuable for organizations working on sovereign AI initiatives that require local compute resources for compliance or low-latency performance.

NVIDIA DGX Cloud Lepton connects our network of global GPU cloud providers with AI developers, said Jensen Huang, founder and CEO of NVIDIA. Together with our NCPs, we’re building a planetary-scale AI factory.

This new platform simplifies and streamlines how developers build, test, and deploy AI applications. It supports multi-cloud and hybrid environments, integrates with NVIDIA’s advanced software tools, and ensures consistent, enterprise-level performance and security.

Built-In Tools for Developers and Providers

DGX Cloud Lepton is tightly integrated with the NVIDIA software ecosystem, including.

• NVIDIA NIM™ and NeMo™ microservices
• NVIDIA Blueprints for prebuilt AI workflows
• NVIDIA Cloud Functions for managing cloud-native AI tasks

For cloud providers, the platform offers advanced GPU management tools that provide real-time diagnostics, automated root-cause analysis, and reduced downtime helping maintain high performance with less manual work.

Key Benefits of DGX Cloud Lepton

• Unified AI development: One platform for building, training, and deploying AI applications.
• Flexible compute options: Choose between marketplace GPU resources or use your own compute clusters.
• Frictionless multi-cloud deployment: Seamlessly deploy across public, private, or hybrid cloud setups.
• Regional sovereignty & performance: Access GPUs in specific regions to comply with local regulations and deliver low-latency AI.
• Predictable, enterprise-grade performance: Reliable, secure, and consistent computing experience.

Introducing NVIDIA Exemplar Clouds

NVIDIA also announced Exemplar Clouds a program to help NCPs improve security, usability, and resilience using NVIDIA’s reference designs, hardware, and operational playbooks.

The Exemplar Clouds will use DGX Cloud Benchmarking, a toolkit for optimizing AI performance and measuring cost-efficiency. Yotta Data Services became the first cloud provider in the Asia-Pacific region to join the initiative.

Availability

• Developers can now sign up for early access to DGX Cloud Lepton.
• More details were shared during Jensen Huang’s COMPUTEX keynote and are available via NVIDIA GTC Taipei.