# zilliztech/deep-searcher

Open Source Deep Research Alternative to Reason and Search on Private Data. Written in Python.

Repository: https://github.com/zilliztech/deep-searcher
Canonical: https://ross.abutalabs.com/products/deep-searcher
Homepage: https://zilliztech.github.io/deep-searcher/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: agent, llm, rag, vector-database, deepseek, agentic-rag, milvus, openai, zilliz, deep-research, deepseek-r1, grok, grok3, reasoning-models, claude, llama4, qwen3
Last push: 2025-11-19T06:04:16+00:00

## Health v2 (maintenance only)
Score: 35/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 53, release rhythm 8, longevity 40
- inputs: {"age_days": 572, "days_push": 287, "days_rel": 560, "gap_med": null, "n_releases_24m": 1}
- flags: prerelease_only
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 8224, forks 797 (observed 2026-08-28T04:10:19.682538+00:00)

## What it is
DeepSearcher is an open-source deep research tool that combines LLMs (DeepSeek, OpenAI, Claude, Qwen, etc.) with vector databases like Milvus to search, reason over, and generate reports from private data. It is a Python application suited for enterprise knowledge management and intelligent Q&A over internal documents.

## Use cases
- run deep research queries over private enterprise documents
- build an intelligent Q&A system on internal knowledge bases
- index local files into a vector database for retrieval
- generate comprehensive research reports from private data
- search internal data securely without sending it to third parties
- compare answers from multiple LLM reasoning models over a private corpus

## When to choose
- you need a self-hosted deep research alternative that keeps data private
- your organization already uses Milvus or Zilliz Cloud for vector storage
- you want agentic RAG with reasoning models over local documents
- you need flexible swapping of LLM and embedding providers

## When to avoid
- you only need simple keyword search without LLM reasoning
- you want a fully managed SaaS research product with no setup
- your data is small enough that a plain LLM chat with file upload suffices
- you need web crawling today, as it is still under development

## Facets
- artifact type: application
- maturity: active
- function: rag, search-engine, vector-database, llm-inference, agent-framework, chatbot
- domain: large-language-models, databases, artificial-intelligence
- platform: python, self-hosted, cli
- tags: deep-research, agentic-rag, milvus, private-data-search, enterprise-knowledge-management, deepseek, document-loader, retrieval-augmented-generation, ai-agents, search, docker

## Member repositories
- zilliztech/deep-searcher (main) score 35

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:19.682538+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:29:31.351763+00:00, confidence not recorded.
  - readme: https://github.com/zilliztech/deep-searcher (fetched 2026-08-28T04:10:19.682538+00:00, sha a64d10b41a45)
  - homepage: https://zilliztech.github.io/deep-searcher/ (fetched 2026-08-29T08:28:07.093319+00:00, sha 2bda6c2df145)
- Data as of 2026-08-30T08:39:29.467469+00:00.