Apheris Launches ApherisFold to Make OpenFold3 Securely Usable in Pharma Environments

Berlin, Germany October 28th, 2025 — Apheris, a leading provider of AI applications for drug discovery, today announced the launch of ApherisFold, an enterprise software product that enables pharmaceutical organizations to securely run, benchmark, and fine-tune the latest co-folding models, including OpenFold3 and Boltz-2, directly within their own IT environments. Apheris develops applications to enable pharma R&D teams to securely use and customize AI models for their research. With ApherisFold, scientists can now bring OpenFold3 into their environments and use it for their research on proprietary data in a matter of hours, without compromising security, intellectual property, or integration needs.

OpenFold3, developed by the OpenFold Consortium and the AlQuraishi Lab at Columbia University, is a class I open-source system according to the Linux Foundation’s framework for open-source software and is capable of predicting structures of proteins bound to small molecules, nucleotides and other proteins. While the model represents a major milestone for structural biology, deploying it securely inside enterprise research environments typically requires extensive engineering, infrastructure setup, and compliance validation. ApherisFold addresses these challenges by providing a production-ready software solution that integrates seamlessly into enterprise R&D workflows.

It enables pharma teams to:

• Run co-folding models locally for secure inference on in-house data

• Benchmark models using public or proprietary datasets to assess domain applicability

• Customize models on internal or third-party data via federated networks

• Visualize results in 3D via an intuitive GUI or access them programmatically via an API

The product builds on Apheris’ established role as the OpenFold Consortium’s partner for federated benchmarking and fine-tuning and its leading role in the Federated OpenFold3 Initiative (powered by the AI Structural Biology Network), which brings together five of the top 20 global pharmaceutical companies to collaboratively fine-tune OpenFold3 on their proprietary datasets.

“OpenFold3 is a robust, open-source foundation model reaching state-of-the-art accuracy for complex structures like protein–ligand and antibody–antigen interactions. Making it usable in enterprise environments requires strong security, compliance, and integration capabilities,” said Robin Röhm, CEO and Co-Founder of Apheris. “With ApherisFold, we’re turning open science into a production-ready product that can be used by enterprises - giving pharma scientists instant, secure access to the world’s most powerful co-folding models for structure and co-folding prediction.”

ApherisFold is available for deployment via AWS CloudFormation or pre-built container images for private cloud and on-prem environments. To learn more or get access, visit the ApherisFold website.

About Apheris

Apheris delivers enterprise-grade AI applications for drug discovery, designed for pharma companies to run and customize securely within their own IT environments. By keeping all data local, organizations maintain full sovereignty over IP-sensitive assets. These local deployments also serve as the foundation for Apheris-hosted federated data networks, where pharma companies collaboratively train and benchmark models on proprietary datasets to unlock more robust and generalizable models for drug discovery. Apheris is trusted by many of the world's largest pharmaceutical companies.

Media contact: press@apheris.com

www.apheris.com