From foundation model to hosted AI solution in minutes

What if you could build a custom AI on your own data with a single API call? Learn how to deploy powerful foundation models in minutes.

#1about 3 minutes

Introducing the IONOS AI Model Hub for easy inference

The IONOS AI Model Hub provides a simple REST API for accessing open-source foundation models and a vector database for RAG.

#2about 1 minute

Exploring the curated open-source foundation models available

The platform offers leading open-source models like Meta Llama 3 for English, Mistral for European languages, and Stable Diffusion XL for image generation.

#3about 7 minutes

How to implement RAG with a single API call

Retrieval-Augmented Generation (RAG) is simplified by abstracting vector database lookups and prompt augmentation into one API request using collection IDs and queries.

#4about 1 minute

Building end-to-end AI solutions in European data centers

Combine the AI Model Hub with IONOS Managed Kubernetes to build and deploy full AI applications within German data centers for data sovereignty.

#5about 3 minutes

Enabling direct GPU access within managed Kubernetes

The NVIDIA GPU Operator will enable direct consumption of GPU resources within IONOS Managed Kubernetes by automatically installing necessary drivers and components.

#6about 3 minutes

Deploying custom inference workloads with NVIDIA NIMs

Use the GPU Operator to request GPUs in a pod spec and deploy NVIDIA Inference Microservices (NIMs) to run custom, containerized AI models on your own infrastructure.