LearningCircuit/local-deep-research

Project overview

A locally hosted research assistant that performs multi-source deep research, builds an encrypted searchable knowledge base, and generates cited reports without telemetry. It supports multiple LLMs and search engines while keeping all usage data in isolated, AES-256 encrypted databases.

Project type

AI Agent · RAG · Model Runtime

Use cases

Knowledge Q&A · Search & Research

Deployment

Refer to project documentation

License

MIT

Best for

• Researchers and general users who require a local-only data boundary with no telemetry or tracking for deep research tasks.
• Users with an AVX-capable CPU who want to run research tasks with the option of free local resources or premium search engines and cloud LLMs.

Key capabilities

• Performs deep, agentic research using multiple LLMs and search engines with proper citations.
• Searches across web, academic papers, and custom sources using 10+ free and premium search engines.
• Downloads research sources, extracts text, indexes it, and makes it searchable within an encrypted library.
• Contains no telemetry, no analytics, and no tracking. Usage metrics stay in the local encrypted database.
• Provides per-user isolated SQLCipher databases encrypted with AES-256.
• Provides an MCP server for AI assistants like Claude Desktop and Claude Code to perform deep research.
• Generates reports with sections and a table of contents, exportable as PDF or Markdown.

Limitations and risks

• The --network host option only works on native Linux.
• The MCP server is designed for local use only via STDIO transport and includes no built-in authentication.

Getting started

• Install via a pip package or pre-built Docker images, which include pre-built SQLCipher wheels. Run the application and open localhost:5000 to begin.

Evidence and sources

• GitHub project description: ~95% on SimpleQA (e.g. Qwen3.6-27B on a 3090). Supports all local and cloud LLMs (llama.cpp, Ollama, Google, ...). 10+ search engines - arXiv, PubMed, your private documents. Ever…
• README: AI research assistant you control. Run locally for privacy, use any LLM and build your own searchable knowledge base. You own your data and see exactly how it works.
• README: AI-powered research assistant for deep, agentic research
• README: Performs deep, agentic research using multiple LLMs and search engines with proper citations
• README: **Option 2: Docker Compose** CPU-only (all platforms): ```bash curl -O https://raw.githubusercontent.com/LearningCircuit/local-deep-research/main/docker-compose.yml && docker comp…

Reviewed at: 2026-07-16T10:26:19Z