GPT Researcher
Autonomous deep-research agent that produces cited reports from web and local sources.
About
GPT Researcher is an open-source autonomous agent for deep research on any task, designed for both web and local research. Planner and execution agents decompose each task into research questions, decision models filter context down to what is actually useful, and the agent produces detailed, factual research reports with citations. The project positions itself as going well beyond RAG, with a full suite of customization options for building tailored, domain-specific research agents.
It is maintained by Tavily's Assaf Elovic, ships on PyPI and Docker with a CLI and Python API, supports any LLM provider, and carries an official site (gptr.dev) plus documentation (docs.gptr.dev). With over 30k GitHub stars it is the reference open implementation of the deep-research agent pattern, popular for market research, due diligence, and literature reviews.
What it can do
- Planner + execution agents that decompose research tasks
- Cited, structured research reports
- Web and local-document research
- Smart context filtering to keep only useful context
- Multi-LLM provider support
- PyPI package, Docker image, and CLI
Setup
- pip install gpt-researcher (or use the Docker image)
- Set OPENAI_API_KEY and TAVILY_API_KEY (other search/LLM providers supported)
- Run a research task via CLI or the Python API
Prerequisites
Python 3.11+; an LLM API key and a search API key (e.g. Tavily).
Use cases
Works with
Trust score — 9/10
Scored 9/10 (verified tier, metrics-v1). Evidence: Verified via official GitHub repo + gptr.dev + docs.gptr.dev, batch277; GitHub 30,112 stars; deep-research agent pattern, planner/executor + cited reports; Trending score 90/100 from researcher assessment. License: Apache-2.0.
★ 30,112 GitHub stars
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.