PrivateGPT

Run a Retrieval-Augmented Generation (RAG) pipeline on your documents, 100% privately, with no data ever leaving your local environment.

PrivateGPT is a production-ready AI project: it delivers a fully private, context-aware LLM application. The core function is a complete RAG pipeline, managing ingestion, embedding, and querying across various document formats (PDF, DOCX, TXT). Crucially, the system ensures 100% data privacy by executing entirely within your environment, even supporting fully local setups with models like Ollama. It provides a robust API that extends the OpenAI API standard, making it a seamless drop-in replacement for existing projects requiring data security and control.

https://privategpt.dev