GPT Researcher is an autonomous agent that conducts deep research on any topic or dataset and produces detailed, cited reports. Give it a question, and it plans sub-questions, searches the web and your own documents in parallel, cross-checks the sources, and writes up the findings with references.
GPT Researcher has grown to more than 30,200 stars on GitHub and is one of the most widely used open-source deep research projects, with active development through October 2026. Its architecture mirrors a human research team: a planner breaks the query into sub-questions, executor agents gather information for each one from web search and local files, and a publisher agent aggregates everything into a coherent report. Unlike a single search-and-summarize pass, it runs many retrievals in parallel, filters for relevance, and keeps citations attached to claims. It works with any major LLM provider, ships an MCP server for integration into agent workflows, and offers both a Python package and a ready-to-use web application.
The open-source version is free to self-host (you pay your own model and search API costs). Hosted plans on the official site add a managed app, higher limits and team features for those who prefer not to run it themselves.
GPT Researcher is for analysts, journalists, students, consultants and product teams who need thorough, source-backed research reports in minutes, and for developers embedding deep-research capabilities into their own AI products.