An autonomous research agent that generates detailed, cited reports by aggregating information from over 20 web or local sources in parallel. It is designed for researchers and developers needing current, long-form research without manual web scraping.
Project overview
It addresses LLM context limits by scraping multiple sites per research query, selecting the most frequent information, and organizing it into comprehensive reports with citations.
Project type
AI Agent · RAG · AI Search
Use cases
Knowledge Q&A · Search & Research · Automation
Deployment
Python environment
License
Apache-2.0
Best for
Researchers and general users looking to generate long-form, cited research reports by aggregating information from over 20 web or local sources in parallel.
Developers needing an autonomous research agent or library that integrates with frameworks like LangGraph and supports custom data sources via MCP.
Key capabilities
Conducts autonomous deep research on web and local documents using planner and execution agents to generate detailed reports with citations.
Conducts research tasks using local documents, including PDF, plain text, CSV, Excel, Markdown, PowerPoint, and Word files.
Executes a tree-like exploration pattern with configurable depth and breadth for comprehensive research.
Connects with specialized data sources such as GitHub repositories, databases, and custom APIs alongside standard web search using Model Context Protocol integration.
Provides multi-agent assistants built with LangGraph and AG2 frameworks where multiple agents with specialized skills collaborate to conduct research.
Automatically generates and embeds AI-created illustrations within research reports using Google's Gemini models.
Offers a lightweight static frontend served by FastAPI and a feature-rich NextJS application for interacting with the research process.
Limitations and risks
Current LLMs have token limitations that are insufficient for generating very long research reports without aggregation or chunking strategies.
Executing deep research incurs an approximate cost of $0.4 per research run due to external API usage.
LLMs trained on outdated information can hallucinate or become irrelevant for current research tasks if not properly grounded.
The project is an experimental application provided 'as-is' without any warranty.
Getting started
Setup is rated as medium difficulty because it requires configuring API keys for LLM and search providers across multiple deployment options. A standard path involves cloning the project, setting up a .env file with API keys, installing dependencies via pip, and starting the server with uvicorn.
Evidence and sources
GitHub project description: An autonomous agent that conducts deep research on any data using any LLM providers
README: 🖥️ Frontend available in lightweight (HTML/CSS/JS) and production-ready (NextJS + Tailwind) versions.
README: pip install gpt-researcher
README: - Objective conclusions for manual research can take weeks, requiring vast resources and time. - LLMs trained on outdated information can hallucinate, becoming irrelevant for curr…
README: GPT Researcher the first open deep research agent designed for both web and local research on any given task. The agent produces detailed, factual, and unbiased research reports w…