Conversational AI is revolutionizing map interactions through the introduction of 'conversational maps,' which allow users to engage with maps using natural language, thus transforming passive exploration into interactive and actionable guidance. The Mapbox Location Agent exemplifies this innovation by integrating a large language model (LLM) with the Mapbox MCP Server to provide users with visual responses via a map-based interface. This system combines LLM reasoning capabilities with geospatial data and routing algorithms to offer seamless, multi-turn conversations for tasks such as trip planning, site selection, and more. Users can bypass traditional multistep processes by simply asking questions in natural language, receiving tailored, spatially aware answers. This technology enables complex workflows like determining optimal travel routes or evaluating neighborhood attributes to be executed with a single prompt. The Mapbox Location Agent demo and the Mapbox MCP Server underscore the potential for smarter, more personalized map experiences, inviting developers to explore and contribute to the evolution of conversational maps.