Local Deep Research

What is Local Deep Research?

Local Deep Research (LDR) is an open-source research assistant you run yourself. You ask a complex question, and it searches the web, academic sources such as arXiv, PubMed and Semantic Scholar, and your own documents, then writes a report with citations. A LangGraph agent strategy lets the model decide what to search and when to synthesize, while quick pipeline modes handle simple fact lookups. Sources can be downloaded into an encrypted library, indexed and searched later alongside the live web. Each user gets an isolated SQLCipher database with AES-256 encryption, and the README states there is no telemetry or analytics. Egress scopes such as Private only and Strict restrict, on a best-effort basis, where a run may send traffic. It works with local models through Ollama or any OpenAI-compatible endpoint, and with cloud providers, and it installs by Docker, Docker Compose or pip. The project is open source and free; you pay only for your own hardware or any cloud model you connect.

Use cases

• Run deep multi-step research with citations
• Search arXiv and PubMed from one assistant
• Query your private documents with AI
• Build a personal searchable knowledge base
• Keep research local and encrypted
• Get quick cited summaries of a question

Who it's for

• Researchers
• Self-hosting enthusiasts
• Students
• Privacy-focused professionals

Why use Local Deep Research

• Running on your own hardware keeps your questions and findings out of third-party research services.
• You choose the model and search engines, so you are not tied to one vendor.
• Citations in each report make it easier to check answers before relying on them.
• Downloaded sources accumulate in a searchable library, so later research can draw on what you already collected.

Local Deep Research features

• LangGraph agent strategy for autonomous research
• Quick Summary mode for answers in 30 seconds to 3 minutes
• Searches web, academic papers and your own documents
• Reports with citations
• Encrypted library for downloaded sources
• Per-user SQLCipher database with AES-256
• Egress scopes: Adaptive, Public only, Private only, Strict
• Local models via Ollama or OpenAI-compatible endpoints
• Docker, Docker Compose and pip installs
• No telemetry, analytics or tracking

Local Deep Research integrations

• Ollama
• SearXNG
• arXiv
• PubMed
• Semantic Scholar
• llama.cpp
• LM Studio
• Paperless-ngx
• Elasticsearch
• Apprise
• Serper

How to get started with Local Deep Research

1. Pick an install route: Docker run, Docker Compose, or pip install local-deep-research.
2. Run Ollama or another OpenAI-compatible model endpoint, plus SearXNG for web search.
3. Open the web interface at http://localhost:5000 once the container or app has started.
4. Enter a research question, choose a research mode, models and sources, and set an egress scope if needed.
5. Read the cited report and save the sources to your encrypted library for later searches.

Local Deep Research limitations

• Self-hosted: you run the app plus Ollama or a compatible endpoint and SearXNG
• Docker --network host works only on native Linux, so Mac, Windows and WSL2 users need Docker Compose
• Needs an x86-64 CPU with AVX (Sandy Bridge or newer) or an ARM64 CPU
• Egress controls are best-effort guard rails, not an air gap, and a local endpoint relaying to a hosted model can still send prompts out

Local Deep Research pricing

Local Deep Research is open source, so you can use and self-host it for free.

Prices can change and may vary by country or exclude tax. Check the official pricing page.

Local Deep Research alternatives

The main Local Deep Research alternatives listed on Debuly are GPT Researcher, Perplexica, Perplexity, MemFree, II-Commons Skills and ChatPDF. The table compares pricing and platforms; order is A–Z.

1. 

   ChatPDF

   Upload a PDF, Word or PowerPoint file and ask questions in any language, with answers linked to the source page.

   Research and searchProductivityWriting
   FreemiumVisit ChatPDF (official site)
2. 

   GPT Researcher

   Open source research agent that plans questions, searches the web or your files, and writes a cited long-form report.

   Research and searchAutomation and agentsDeveloper platforms
   Open sourceVisit GPT Researcher (official site)
3. IC

   II-Commons Skills

   A command-line tool and agent skill for fast retrieval of arXiv, PubMed, PMC and US policy documents from daily-updated corpora.

   Research and searchDeveloper platforms
   Visit II-Commons Skills (official site)
4. 

   MemFree

   Open-source hybrid AI search engine that answers from your own knowledge base and the web, plus an AI page generator for React and Tailwind UIs.

   Research and searchDeveloper platformsCoding
   Open sourceVisit MemFree (official site)
5. Pe

   Perplexica

   Open-source, self-hosted AI answering engine (now Vane) that pairs SearxNG web search with local or cloud models and cites its sources.

   Research and searchDeveloper platforms
   Open sourceVisit Perplexica (official site)
6. 

   Perplexity

   AI answer engine that searches the web, cites sources inline and offers deep research, multi-step tasks and a choice of top models on paid plans.

   Research and searchChat assistantsProductivity
   FreemiumVisit Perplexity (official site)

Local Deep Research FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: none collected yet, so a monogram is shown.
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.