A dedicated server that wraps Google's Gemini AI models in a Model Context Protocol (MCP) interface, allowing other LLMs and MCP-compatible systems to access Gemini's capabilities like content generation, function calling, chat, and file handling through standardized tools.