ConechoAI/openai-websearch-mcp

Overview

An MCP server that wraps OpenAI's web search functionality, enabling AI agents to perform web queries through the Model Context Protocol. Built in Python, it acts as a bridge between MCP-compatible clients and OpenAI's web search API.

Best for

Best for
Developers creating MCP-compatible AI agents that need web search capabilities

Use cases

• Integrating web search into MCP-based AI agents
• Building search-enhanced chat or assistant tools via MCP
• Prototyping AI workflows that require real-time web data

How to use

Install

npx @modelcontextprotocol/inspector uvx openai-websearch-mcp

Tools exposed

• reasoning_effort
• search_context_size
• user_location
• gpt-4o
• gpt-4o-mini
• gpt-5-mini
• gpt-5
• gpt-5-nano
• o4-mini
• OPENAI_API_KEY
• OPENAI_DEFAULT_MODEL

Tested with

Claude Desktop, Claude Code, Cursor, ChatGPT

Notes

An MCP server that wraps OpenAI’s web search functionality, enabling AI agents to perform web queries through the Model Context Protocol. Built in Python, it acts as a bridge between MCP-compatible clients and OpenAI’s web search API.

89 stars on GitHub. Last updated 2025-09-12. Licensed MIT.

Use cases

• Integrating web search into MCP-based AI agents
• Building search-enhanced chat or assistant tools via MCP
• Prototyping AI workflows that require real-time web data

Pros

• Leverages OpenAI’s reliable web search API
• Simple MCP interface for easy integration with compatible clients
• Python implementation is straightforward to deploy and extend

Cons

• Requires a valid OpenAI API key with web search access
• Dependent on OpenAI’s API availability and pricing
• Limited to the scope and quality of OpenAI’s web search results

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.