GPT Research

Provides research capabilities by querying GPT-5.1's web search to return cited results.

python

GitHub Stars

python

Language

4 months ago

First Indexed

3 weeks ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "testusuke-gpt-research-mcp": {
      "command": "uv",
      "args": [
        "run",
        "--directory",
        "/path/to/gpt-research-mcp",
        "main.py"
      ],
      "env": {
        "OPENAI_API_KEY": "sk-...",
        "OPENAI_BASE_URL": "https://api.openai.com/v1/",
        "LANGFUSE_BASE_URL": "https://cloud.langfuse.com",
        "LANGFUSE_PUBLIC_KEY": "pk-lf-...",
        "LANGFUSE_SECRET_KEY": "sk-lf-..."
      }
    }
  }
}

You run a research MCP server that uses GPT-5.1's Responses API with an integrated Web Search tool to perform queries and return results with citations. This server makes it easy to perform structured, source-backed research from an MCP client and to run the server locally or register it with an MCP platform for seamless access.

How to use

You can attack research topics by calling the research tool from any MCP client that supports tools. Use the function name research with your query as a string. The server will return the results generated by GPT-5.1, including cited information from sources.

How to install

Prerequisites: You need Python 3.12 or newer. You also need the uv package manager and an OpenAI API key.

# 1. Clone the MCP server repository
git clone https://github.com/your-username/gpt-research-mcp.git
cd gpt-research-mcp

# 2. Install dependencies using the uv package manager
uv sync

# 3. Set your OpenAI API key
export OPENAI_API_KEY="sk-..."

# Optional: set a custom OpenAI base URL if not using the default endpoint
export OPENAI_BASE_URL="https://your-custom-endpoint.com/v1"

# Optional: enable LangFuse tracing (set all three vars to activate)
export LANGFUSE_SECRET_KEY="sk-lf-..."
export LANGFUSE_PUBLIC_KEY="pk-lf-..."
export LANGFUSE_BASE_URL="https://cloud.langfuse.com"  # EU region

Register the MCP server with Claude Code. You can add it under your MCP list so you can run it from the Claude client as an MCP. Use the following command to register the server definition, pointing to the local runtime directory and entry file.

claude mcp add gpt-research -- uv run --directory /path/to/gpt-research-mcp main.py

You can also store the MCP configuration directly in your Claude settings, under the mcpServers section, using the following JSON structure.

## Additional setup details

If you prefer to store the MCP server configuration locally, include: OPENAI\_API\_KEY and OPENAI\_BASE\_URL in the environment settings for the server so that the runtime can access the OpenAI API.

{ "mcpServers": { "gpt_research": { "command": "uv", "args": ["run", "--directory", "/path/to/gpt-research-mcp", "main.py"], "env": { "OPENAI_API_KEY": "sk-...", "OPENAI_BASE_URL": "https://api.openai.com/v1/" } } } }


`OPENAI_BASE_URL` is optional. If you are not using a custom endpoint, you can omit it.

Usage in MCP clients

To use the tool from an MCP client (for example Claude Desktop), call the research tool with your query string. Example in a client environment:

research("Tell me about AI trends in 2024")

Notes

The server relies on GPT-5.1 with a built‑in web search capability to provide cited results. Ensure your OpenAI API key is valid and has access to the web search features if required by your plan.

Available tools

research

Fetches and analyzes a topic using GPT-5.1's web search capabilities and returns results with citations.

Built by

VeilStrat

AI signals for GTM teams