GPTR LogoGPT Researcher

/developers

Developer hub

GPT Researcher ships across runtimes and clients: a Python SDK, a Docker image, a Model Context Protocol server, and an OpenAPI 3.1 REST spec. Pick the integration path that matches your stack.

Python

pip install gpt-researcher

from gpt_researcher import GPTResearcher
import asyncio

async def main():
    r = GPTResearcher(query="your topic", report_type="research_report")
    await r.conduct_research()
    print(await r.write_report())

asyncio.run(main())
Reference

MCP (Streamable HTTP)

claude_desktop_config.json or mcp-remote

{
  "mcpServers": {
    "gptr-public": {
      "url": "https://gptr.dev/api/mcp"
    }
  }
}
Reference

Self-hosted MCP (full)

git clone https://github.com/assafelovic/gptr-mcp

git clone https://github.com/assafelovic/gptr-mcp
cd gptr-mcp
pip install -r requirements.txt
python server.py  # stdio
# or:
docker compose up -d  # SSE on :8000/sse
Reference

REST API

OpenAPI 3.1 - any language

curl -X POST https://your-self-hosted-server/research \
  -H "Content-Type: application/json" \
  -d '{"query":"What are the strategic risks for NVIDIA?","report_type":"research_report"}'
Reference

Verified integrations

GPT Researcher integrates as a tool / MCP server with the major AI agent platforms. Each link below points to the canonical install instructions.

Agent-discovery endpoints

These files let any AI agent discover, understand, and connect to GPT Researcher without a human in the loop. They are kept in sync with the live MCP server at /api/mcp.