SERVICES YOU CAN LAUNCH WITH THE RAFAY PLATFORM

NVIDIA NIM-Powered Model Marketplace as a Service, Delivered by Rafay

The NIM-Powered Model Marketplace as a Service, delivered by Rafay, enables telecommunications providers and sovereign cloud operators to transform NVIDIA GPU infrastructure into fully operational AI marketplaces.

NVIDIA NIM (NVIDIA Inference Microservices) are containerized, performance-optimized microservices that run AI models on NVIDIA GPUs and expose industry-standard APIs. The Rafay Platform is an infrastructure orchestration and automation platform that enables self-service consumption of compute and AI services.

Built on NVIDIA NIM and orchestrated by Rafay, this solution allows telcos to launch branded AI marketplaces where enterprises can select, deploy, and consume AI services instantly using self-service, with full governance, sovereignty, and monetization controls.

Static model catalogs lack operational readiness, custom integrations delay launches, and enterprises need sovereign, compliant services.

Rafay and NVIDIA NIM solve this by enabling telcos to stand up governed, production-ready marketplaces in weeks, connecting compute, models, and blueprints in a unified platform.

  • Models-as-a-Service: Expose NVIDIA NIM-powered models as governed API endpoints with tenant-level usage tracking‍
  • Blueprints-as-a-Service: Offer NVIDIA-curated blueprints (fraud detection, medical imaging, data flywheels) deployable in one click‍
  • Compute-as-a-Service: Provide on-demand provisioning of Kubernetes clusters, virtual machines, and GPU slices
hero image of Rafay AI inference as a service page

Operationalize AI Services Across Compute and Models

Rafay provides the orchestration and automation layer needed to operationalize AI service delivery for telcos, enterprises, and sovereign cloud providers.

Unified Marketplace Experience

White-labeled portals allow enterprises to browse, deploy, and consume AI services instantly.

Governed Service Delivery

Built-in usage tracking, compliance, and sovereignty ensure every deployment meets enterprise and regulatory standards.

API Compatibility

OpenAI-compatible APIs enable applications to access AI models through standardized endpoints.

Rapid Deployment Framework

Telcos can stand up fully branded AI marketplaces in weeks, drastically reducing time to production.

Why Use Rafay for NVIDIA NIM-Powered AI Services?

Launch operationalized, NIM-powered AI marketplaces in weeks instead of months

Move Beyond GPU Capacity With Differentiated AI Services

Support Sovereign and Regulated AI Deployments

Maintain Control Over AI Infrastructure and Operations

Give Customers a Branded Self-Service Experience

NIM-Powered Model Marketplace FAQs

What is NVIDIA NIM?

NVIDIA NIM (NVIDIA Inference Microservices) provides packaged inference capabilities for AI models. Rafay supports NVIDIA NIM alongside other AI technologies. The platform allows organizations to run NIM-powered models on GPU systems and make them available as governed services through the Rafay Platform.

How does Rafay support NVIDIA NIM for LLMs?

Rafay enables organizations to provision and manage endpoints for LLMs and other AI workloads running on GPUs. Providers can use NIM-powered models as part of a catalog of AI services. Self-service access, tenant-level governance, usage tracking, and infrastructure orchestration can be managed through the Rafay Platform.

What can providers offer through a NIM-powered AI marketplace?

Providers can offer NVIDIA NIM-powered Models-as-a-Service, NVIDIA Blueprints, and Compute-as-a-Service with Kubernetes clusters, virtual machines, and GPU slices—all through a catalog that extends beyond GPU capacity alone.

How are NIM inference endpoints delivered to customers?

Rafay exposes NIM-powered models as governed API endpoints. Customers can deploy available services through a self-service experience and use the resulting endpoint to send requests to the model and receive a response, while providers maintain tenant-level governance and usage visibility.

Can Rafay support self-hosted NIM services?

Yes. The Rafay Platform can be consumed as SaaS or deployed in self-hosted environments, and Rafay supports NIM and AI service delivery across private, sovereign, and air-gapped environments. This allows providers to operate AI services while maintaining control over their systems and data.

What are the prerequisites to get started with NIM on Rafay?

Rafay's NIM documentation lists a shared host Kubernetes cluster with GPUs as a prerequisite for its NIM Microservices template. Administrators configure the required service profile and model settings, including an NGC API key, before making the service available to end users through the Rafay catalog. Detailed configuration requirements are available in the Rafay product documentation.

Does Rafay provide the NVIDIA GPUs or AI models?

No. Rafay provides the orchestration, automation, governance, and service-delivery layer for AI infrastructure. NVIDIA and other providers supply the underlying GPU hardware and AI technologies. Rafay enables providers to turn those resources into governed, self-service services that customers can consume.

Can NIM services be accessed through APIs?

Yes. Rafay supports NIM-powered models as API endpoints, allowing applications and users to interact programmatically with deployed inference services. Rafay also provides APIs and other programmatic interfaces that providers can use to integrate platform capabilities into their existing portals and service-delivery environments.

Start a conversation with Rafay

Talk with Rafay experts to assess your infrastructure, explore your use cases, and see how teams like yours operationalize AI/ML and cloud-native initiatives with self-service and governance built in.

“We are able to deliver new, innovative products and services to the global market faster and manage them cost-effectively with Rafay.”

Joe Vaughan
Joe Vaughan
Chief Technology Officer
,
MoneyGram
Whitepaper

Hybrid Cloud Meets Kubernetes

Learn how to Streamline Kubernetes Ops in Hybrid Clouds with AWS & Rafay