openai-websearch-mcp is a Python-based MCP server that exposes the OpenAI web_search tool to enable AI assistants to perform up-to-date web searches with built-in reasoning. It supports multiple OpenAI reasoning models (gpt-5, gpt-5-mini, gpt-5-nano, o3, o4-mini) and offers smart reasoning_effort defaults tailored to the use case. The server supports a multi-mode search approach: fast iterations with smaller models (gpt-5-mini) and deep research with more capable models (gpt-5). Localized results can be produced via user_location for location-based search customization. It provides complete parameter documentation for easy integration and supports flexible configuration through environment variables to simplify deployment. The README also describes quick-start installation for Claude Desktop, Cursor, and local development, along with debugging tooling including the MCP Inspector, illustrating practical deployment and troubleshooting steps. In short, openai-websearch-mcp combines reasoning-enabled web search with MCP compatibility to power AI assistants requiring current information and thoughtful analysis.