# GPT Researcher

GPT Researcher is an AI research agent that searches more than twenty sources to produce detailed, factual, well-cited reports, saving hours of manual searching and summarising. It is open source and self-hosted, installable on Docker Hub.

Canonical page: https://ourlocalhost.com/app/6354/gpt-researcher

## Description

GPT Researcher is an AI research agent that searches more than twenty sources to produce detailed, factual, well-cited reports, saving hours of manual searching and summarising.

## Platform availability

Available on 0 of 10 platforms.

## Health score

79/100 (Good), based on 5 of 5 factors.
- Maintained: 100 (last commit 12 days ago)
- Popular: 90 (30,786 GitHub stars)
- Easy to install: 25 (0 install platforms, install notes, Docker image)
- Light to run: 60 (2 GB minimum RAM, runs on ARM)
- Security check: 89 (no serious findings, 1 minor, CI, tests)

## Links

- GitHub: https://github.com/assafelovic/gpt-researcher
- Docker Hub: https://hub.docker.com/r/elestio/gpt-researcher
- Official website: https://gptr.dev
- Documentation: https://docs.gptr.dev

## Common questions

### Why does GPT Researcher report that model: gpt-4 does not exist?

It indicates that your OpenAI account does not yet have permission to use gpt-4.

### What environment variables are required for the gptr-mcp server?

Set OPENAI_API_KEY and TAVILY_API_KEY in the MCP server environment and run the local server.py file.

### Can GPT Researcher research local documents instead of only the web?

Yes, it can research PDFs, Word files, spreadsheets, and Markdown. These files can be used alone or combined with web sources.

### Do I need an embeddings provider for the default source filtering?

No, it falls back to local keyword ranking, so no key or embeddings provider is required.

---
Source: Our Local Host (https://ourlocalhost.com/app/6354/gpt-researcher), the self-hosted and open-source app directory.
