Frequently Asked Questions

OpenMed is an open-source, local-first healthcare AI solution that processes clinical text entirely on-device. It provides entity extraction, PII de-identification, and access to 1,000+ specialized medical models without sending patient data to the cloud. Built with privacy and compliance as core principles.

OpenMed never sends patient data to external servers. All processing happens locally on the device or within your network. It includes HIPAA-compliant PII de-identification with smart entity merging, format-preserving fakes, and cryptographic hashing options. The solution is designed for healthcare compliance with zero data leakage.

OpenMed runs on Linux, macOS, Windows (CPU/CUDA), and Apple Silicon (MLX). It includes native iOS/macOS apps via OpenMedKit and can be deployed as a Python API, Dockerized REST service, or batch processing pipeline. Works offline and in air-gapped environments.

OpenMed provides access to 1,000+ specialized biomedical and clinical NER models covering diseases, drugs, anatomy, genes, and more. Models are curated from public datasets and many outperform proprietary solutions. The model registry is continuously expanding with domain-adapted transformers.

OpenMed supports 12 languages for PII extraction and de-identification: English, Spanish, French, German, Italian, Dutch, Hindi, Arabic, Japanese, Turkish, Portuguese, and Telugu. This includes 247 PII checkpoints with language-specific clinical ID providers.

Installation is simple: pip install "openmed[hf]" for core functionality, pip install "openmed[mlx]" for Apple Silicon acceleration. See our documentation for quick start guides, API examples, and deployment options.

OpenMed is released under the Apache License 2.0, which allows commercial use, modification, and distribution. There are no usage restrictions or vendor lock-in. You can use it in healthcare systems, research projects, or commercial products without royalties.

Unlike cloud medical APIs, OpenMed runs entirely on your hardware with no data transmission costs. It offers more specialized medical models, true offline operation, and zero vendor lock-in. With Apple MLX acceleration, it provides 24-33x faster performance than cloud solutions for critical tasks.

Yes! OpenMedKit provides a native Swift framework for iOS and macOS apps. It enables on-device PII detection and clinical extraction with Apple MLX acceleration. Perfect for mobile health apps requiring HIPAA compliance without network dependencies.

Support is available through our GitHub repository for bug reports and feature requests. Comprehensive documentation is available at openmed.life/docs. Join our community on X/Twitter and LinkedIn for updates and discussions.