Supporting competitiveness and innovation potential of SMEs

Overview

Funding for SMEs and startups specializing in generative AI, software development, or data processing to conduct innovation studies that develop or customize generative AI models using large-scale European high-performance computing (HPC) resources. The focus is on addressing business challenges where existing AI models are insufficient due to technical, licensing, or other constraints.

Who Can Apply

SMEs and startups with expertise in generative AI, professional software development, or data processing. Applicants may only participate in one of the two sub-project types under the FFplus call (this call or the business experiments call). The applicant must serve as the main participant.

Funding Details

Details on grant amounts, total budget, or funding rates are not specified in the provided information.

What Is Funded

Innovation studies that:

• Develop or customize generative AI models (e.g., foundation or large language models) using European HPC resources (pre-exascale or exascale supercomputers).
• Address business problems where existing AI models are inadequate, with justification for why a new model is necessary.
• Include a detailed machine learning lifecycle: data preparation, model development/engineering, and evaluation.
• Demonstrate availability of suitable training datasets and define model characteristics (type, size, hyperparameters, architecture).
• Establish performance metrics, reproducibility plans, and risk mitigation strategies aligned with EU trustworthy AI guidelines.
• Produce a data management plan covering data access, usage, sharing, retention, and compliance with FAIR principles.
• Support the generation of success stories, including multimedia content, highlighting business benefits, technical challenges, and societal or environmental impact.

Technology / Maturity Requirements

• Proposals must justify the use of large-scale HPC resources for model development.
• Applicants must demonstrate technical readiness to execute the proposed ML lifecycle, including data preparation, model training, and evaluation.
• No explicit Technology Readiness Level (TRL) range is specified.

Key Dates

• Open date: 2025-12-02T23:00:00Z
• Deadline: 2026-02-24T23:00:00Z

Application Process (High-Level)

Details on the application process (e.g., single-stage or multi-stage, submission portal) are not specified in the provided information.

Post-Award Obligations

• Deliver an intermediate report on results and potential impact by the end of the 7th month, which will determine eligibility for a potential extension under a subsequent call.
• Support the generation of success stories, including multimedia content, for publication.
• Comply with reporting requirements related to business impact, technical outcomes, and societal/environmental effects.

Geographic Scope & Language

• Eligible countries: Not explicitly specified (assumed to align with EU participation).
• Application language: English.

Notes

• Participation in this call is mutually exclusive with the FFplus business experiments call (Identifier FFplus_Call-2-Type-1).
• Proposals must justify why existing generative AI models cannot address the business problem (e.g., due to licensing, few-shot prompting limitations).
• Applicants must demonstrate the availability of training datasets and provide a detailed data management plan.
• Emphasis on reproducibility, performance metrics, and risk mitigation in line with EU trustworthy AI guidelines.