Deploy a model in production

Once you have trained your model, you can deploy it in production. This section provides several guides in how to deploy your model in different environments.

 Overview

Start here to get an overview of the different deployment options.

Deployment options
 Deploy as serverless

Deploy your model in the platform using the serverless option, using a shared serverless environment.

Deploy a model on the Inference platform using the Dashboard
 Deploy persistently

Deploy your model in the platform using a dedicated deployment and a load balancer.

Deploy a model on dedicated resources using the Dashboard
 Deploy in your cloud

Deploy your model in your cloud using the provided Docker image.

Deploy a model on your own cloud resources
 Deploy in the EOSC EU Node

Deploy your model in the resources provided by the EOSC EU Node.

Deploy a model on the EOSC EU Node
 Deploy external models

Deploy models from external marketplaces (BioImage Model Zoo).

Deploy an external model
 Deploy your own LLM

Deploy your own LLM from a selection of open-source models (DeepSeek, Qwen, LLama, etc), using vLLM and Open-WebUI.

Deploy your own LLM chatbot