OptiMUS is a Large Language Model (LLM)-based agent designed to formulate and solve (mixed integer) linear programming problems from natural language descriptions. It addresses the challenge of optimization modeling by automating the process, which traditionally requires expert knowledge. OptiMUS can develop mathematical models, write and debug solver code, evaluate solutions, and improve its model and code based on these evaluations. The system utilizes a modular structure to handle long problem descriptions and complex data without requiring excessively long prompts. Experiments demonstrate that OptiMUS outperforms existing state-of-the-art methods on both easy and hard datasets, including a new dataset called NLP4LP, which features long and complex problems. The paper also introduces NLP4LP, an open-source dataset of 67 complex optimization problems, and discusses the contributions, methodology, and experimental results of OptiMUS.OptiMUS is a Large Language Model (LLM)-based agent designed to formulate and solve (mixed integer) linear programming problems from natural language descriptions. It addresses the challenge of optimization modeling by automating the process, which traditionally requires expert knowledge. OptiMUS can develop mathematical models, write and debug solver code, evaluate solutions, and improve its model and code based on these evaluations. The system utilizes a modular structure to handle long problem descriptions and complex data without requiring excessively long prompts. Experiments demonstrate that OptiMUS outperforms existing state-of-the-art methods on both easy and hard datasets, including a new dataset called NLP4LP, which features long and complex problems. The paper also introduces NLP4LP, an open-source dataset of 67 complex optimization problems, and discusses the contributions, methodology, and experimental results of OptiMUS.