Local Deep Research is a self-hosted research agent that performs comprehensive investigations using LLMs. It supports both local models (llama.cpp, Ollama, Mistral) and cloud providers (OpenAI, Anthropic, Google) with ~95% accuracy on SimpleQA benchmarks. The system incorporates RAG with encrypted SQLite storage to search across academic sources (arXiv, PubMed), web search engines, and private documents. It provides a Docker-based deployment with end-to-end encryption for secure, private research workflows.