The MCP Solver is a Model Context Protocol (MCP) server that exposes constraint solving, SAT, SMT, and ASP capabilities to Large Language Models. It integrates five backends—MiniZinc, PySAT (including MaxSAT), Z3, and Clingo—to enable interactive modeling and solving via the MCP. The server supports multiple problem paradigms: MiniZinc constraint models, SAT/CNF, MaxSAT optimization, SMT formulas, and Answer Set Programs. Users interact through a compact command surface to build and manipulate models, including clear_model, add_item, delete_item, replace_item, get_model, and solve_model. For MiniZinc, solutions can be retrieved via get_solution. The MCP Test Client provides a development, experimentation, and diagnostic workflow, requiring an API key for LLM providers and supporting Anthropic, OpenAI, Gemini, OpenRouter, or local models. Backend-specific installation and configuration details are provided, with quick-start guidance using the uv package manager. The README also points to a research paper describing the architecture as a bridge between language models and symbolic solvers through the Model Context Protocol.