OpenAI WebSearch MCP Server

What is OpenAI WebSearch MCP Server about?

Enables AI assistants to perform web searches using OpenAI's websearch endpoint, delivering fresh information that may not be present in the model's training data.

How to use OpenAI WebSearch MCP Server?

1. Install the package (one‑click via uvx or standard pip).
2. Add the server configuration to your Claude, Zed, or other MCP‑compatible client settings, supplying your OpenAI API key.
3. Invoke the web_search tool from the assistant, providing required arguments such as type, search_context_size, and optional user_location.
4. Use the MCP inspector (npx @modelcontextprotocol/inspector) to debug if needed.

Key Features

• web_search tool with required arguments (type, search_context_size) and optional user location metadata.
• Adjustable context size (low, medium (default), high).
• One‑click installation via uvx (OPENAI_API_KEY=sk-xxxx uv run --with uv --with openai-websearch-mcp openai-websearch-mcp-install).
• Manual installation options for both uvx and pip.
• Compatibility with Claude.app, Zed editor, and any MCP‑compatible client.
• Debug support through the MCP inspector.

Use Cases

• Real‑time factual retrieval in chatbots and virtual assistants.
• Supplying up‑to‑date references or citations during a conversation.
• Enhancing knowledge‑base queries with the latest web information.
• Context‑aware searches based on user location (city, country, region, timezone).

FAQ

Q: Do I need an OpenAI API key? A: Yes, the server requires a valid OPENAI_API_KEY environment variable.

Q: Which editors or platforms are supported? A: Claude.app, Zed editor, and any client that implements the Model Context Protocol.

Q: How do I debug the server? A: Use the MCP inspector, e.g., npx @modelcontextprotocol/inspector uvx openai-websearch-mcp.

Q: Can I customize the search context size? A: Yes, set search_context_size to low, medium, or high when calling the tool.

Q: Is user location required? A: No, it is optional but can improve search relevance when provided.

OpenAI WebSearch MCP Server

This MCP server provides access to OpenAI's websearch functionality through the Model Context Protocol. It allows AI assistants to search the web during conversations with users, providing up-to-date information that may not be available in the assistant's training data. The server can be installed and configured for use with Claude.app or Zed editor.

One click installation & Configuration

Claude

!!Can using this command auto update configure file（Recommend）

OPENAI_API_KEY=sk-xxxx uv run --with uv --with openai-websearch-mcp openai-websearch-mcp-install

sk-xxxx is your API key. You can get it from openai's open platform

Cursor

Conming soon

Windsurf

Conming soon

Available Tools

• web_search - Call openai websearch as tool.
  • Required arguments:
    • type (string): web_search_preview
    • search_context_size (string): High level guidance for the amount of context window space to use for the search. One of low, medium, or high. medium is the default.
    • user_location (object or null)
      • type (string): The type of location > approximation. Always approximate.
      • city (string): Free text input for the city of the user, e.g. San Francisco.
      • country (string): The two-letter ISO country code of the user, e.g. US.
      • region (string): Free text input for the region of the user, e.g. California.
      • timezone (string): The IANA timezone of the user, e.g. America/Los_Angeles.

Manual installation and configuration

Please make sure uvx is installed before installation

Add to your Claude settings:

1、Using uvx

"mcpServers": {
  "openai-websearch-mcp": {
    "command": "uvx",
    "args": ["openai-websearch-mcp"],
    "env": {
        "OPENAI_API_KEY": "your-api-key-here"
    }
  }
}

2、Using pip installation

1）install openai-websearch-mcp via pip:

pip install openai-websearch-mcp

2）modify your Claude settings

"mcpServers": {
  "openai-websearch-mcp": {
    "command": "python",
    "args": ["-m", "openai_websearch_mcp"],
    "env": {
        "OPENAI_API_KEY": "your-api-key-here"
    }
  }
}

Configure for Zed

Add to your Zed settings.json:

Using uvx

"context_servers": [
  "openai-websearch-mcp": {
    "command": "uvx",
    "args": ["openai-websearch-mcp"],
    "env": {
        "OPENAI_API_KEY": "your-api-key-here"
    }
  }
],

Using pip installation

"context_servers": {
  "openai-websearch-mcp": {
    "command": "python",
    "args": ["-m", "openai_websearch_mcp"],
    "env": {
        "OPENAI_API_KEY": "your-api-key-here"
    }
  }
},

Debugging

You can use the MCP inspector to debug the server. For uvx installations:

npx @modelcontextprotocol/inspector uvx openai-websearch-mcp