Skip to main content
Nebius AI Cloud offers NVIDIA Blueprint for virtual screening as a Standalone Applications. The application is available in the eu-north1 and us-central1 regions.

​
About the Blueprint

NVIDIA Blueprints are pretrained end-to-end reference AI workflows designed for common enterprise use cases. The NVIDIA Blueprint for generative virtual screening guides researchers through an AI-driven pipeline that designs optimized small molecules efficiently. It orchestrates a suite of GPU-accelerated NIM microservices (including OpenFold2, MSA-Search, GenMol, and DiffDock) to predict protein structures, generate drug-like molecules, and model their docking without predefined binding pockets. This workflow significantly accelerates hit identification and lead optimization, enabling smarter, faster drug discovery cycles. Within the deployment, we also deploy JupyterHub along with the Python notebook sample to make experimentation with the Blueprint more accessible. For more information about deploying the Blueprint and connecting to it, see Deploying NVIDIA Blueprint for virtual screening in Nebius AI Cloud.

​
Terms and conditions

This application is provided under a bring-your-own-license model. By deploying this application, you represent and warrant that you hold, and will maintain throughout the use of the application, a valid and sufficient NVIDIA AI Enterprise License. Use of the application without such license is strictly prohibited. If you do not currently possess the required license, please refer to How to get licenses for NVIDIA Blueprint for virtual screening for available options.

​
Getting started

To use NVIDIA Blueprint for virtual screening, get the licenses and deploy it. After that, you can access it via JupyterHub to run the sample notebook.