GPT Researcher

GPT Researcher is an open deep-research agent for developers, researchers, and teams who need detailed, factual reports instead of doing manual web research or relying on a single LLM's hallucination-prone answers. It works on both web and local documents for any given research task.

Planner and execution agents break a research task into specific questions; a crawler agent gathers information for each question, summarizing and source-tracking every resource. A publisher agent then filters and aggregates the summaries into a final report. Recursive deep research explores a topic in both breadth and depth, and research runs in parallel to keep it fast and reduce misinformation.

It generates detailed reports exceeding 2,000 words that aggregate over 20 sources, with smart image scraping and filtering, AI-generated inline images via Google Gemini, JavaScript-enabled web scraping, and memory maintained throughout a research session. Reports can be exported to PDF, Word, and other formats, and it ships with both a lightweight HTML/CSS/JS frontend and a production-ready NextJS and Tailwind frontend. It can also be installed as a Claude Skill.

GPT Researcher is open source under the Apache-2.0 license, installable via pip, Python, or Docker, and requires API keys such as an OpenAI key and a Tavily key to run. It has more than 30.5k GitHub stars and 4.2k forks.

Use cases

• Researching why a stock is moving
• Building domain-specific research agents