Mederi is Latin for "to heal." OpenMederiAI is a community-driven initiative, built by the community, for the community. We develop and openly share AI methods, models, and benchmarks for healthcare so that anyone can reproduce, use, and improve them.

• Computational Pathology: histopathology image analysis, foundation models, explainable AI
• Clinical AI: time-series modeling, robustness, and calibration
• Biomedical NLP: retrieval-augmented generation for medical text
• Genomics and Drug Discovery: sequence modeling and molecular property prediction

• Community-driven: built in the open, with contributions welcome from researchers, clinicians, and developers.
• Open: code, weights, and evaluation scripts are public where licenses allow.
• Reproducible: documented environments and instructions to rerun every result.
• Rigorous: careful data splits, leakage checks, and transparent reporting of limitations.

For research use only. Not for clinical diagnosis or treatment.

OpenMederiAI grows through community contributions. Open an issue, submit a pull request, start a discussion to collaborate.