AI Initiatives

From advancing trustworthy AI benchmarks to enabling secure model sharing, fostering high-performance computing access, and supporting industry collaborations, the Helmholtz Association and its platforms are building a vibrant AI ecosystem across Germany. These initiatives bring together researchers, companies, and policymakers to address complex challenges in health, energy, climate, and beyond. Whether through community-driven benchmarking calls, cloud-based AI model deployment, cutting-edge HPC gateways, or national AI service centers, Helmholtz is creating the tools, expertise, and infrastructure needed to accelerate AI innovation and translate it into real-world impact. Find out more about these initiatives below.

Unlock Call - Benchmarking Call:

The UNLOCK 2025 call, supported by the Helmholtz Association and the hyperUNLOCK Initiative, seeks to foster a community of practice around benchmarking in AI for science. It aims to support the development of systematic, cross-domain, and multimodal benchmarking datasets that can drive trustworthy, reproducible AI research in Helmholtz-relevant domains such as health, energy, and climate.

Learn more

Who is it relevant for?

This initiative is relevant for AI researchers, data scientists, and domain experts across the Helmholtz Association and beyond, particularly those working with scientific data and interested in building or using benchmarking datasets to evaluate and improve AI models.

Helmholtz Imaging Model Zoo:

The Helmholtz Model Zoo (HMZ) is a cloud-based platform that allows researchers across the Helmholtz Association to easily share, run, and integrate deep learning models via a web interface or REST API. It simplifies model deployment and inference while ensuring data stays securely within the Helmholtz infrastructure.

Learn more

Who is it relevant for?

This initiative is relevant for Helmholtz researchers and AI practitioners who want to publish, test, or apply deep learning models in scientific workflows, particularly those looking for low-barrier, secure, and scalable ways to integrate AI into their research.

HPC gateway:

The HPC Gateway initiative provides companies with access to Helmholtz’s high-performance computing infrastructure, vast datasets, and AI expertise to collaboratively solve AI-related challenges. Through joint projects, it supports knowledge transfer between research and industry while enabling scalable AI applications.

Participating centers: DESY, Hereon, Max Delbrück Center, Helmtolty-Zentrum Berlin, HZDR, FZJ, GSI, KIT, HMGU

Learn more

Who is it relevant for?

This initiative is relevant for companies, especially those working on AI, digital twins, or data-driven innovation, seeking access to advanced computing resources, expert support, and collaborative opportunities with leading research centers in Germany.

HPC Gateway builds on experience from previous projects such as the Smart Data Innovations Lab (SDIL) or the AI service centers and aims to implement such collaborations throughout Germany.

KI Servicezentren (AI Service centers):

The AI Service Centers initiative, launched on September 3, 2021, aims to significantly expand access to powerful AI computing infrastructure in Germany. These centers conduct cutting-edge AI research and offer user-friendly services to transfer AI into practical applications, fostering collaboration between science and industry, especially for SMEs.

Participating centers: hessian AI Service Center, AI Service Center Berlin Brandenburg, KISSKI: A Highly Available AI Service Center for Sensitive and Critical Infrastructures, Service Center for Artificial Intelligence

Learn more

Who is it relevant for?

It is relevant for researchers, businesses, and especially small and medium-sized enterprises (SMEs) seeking access to high-performance AI resources, expertise, and support in developing, understanding, and integrating AI solutions into their operations.

Stay alert for all upcoming opportunities!

Stay up to date with all things AI in science across the Helmholtz Association.

Subscribe