Software effort estimation is the process of estimating the amount of time, effort, and money required to develop and manage software. It has an important role in software project planning and management. It impacts activities such as bidding, budgeting, and resource allocation. However, it is not easy to find a model that fits a project to help in its success. Over the past five decades, a wide range of models and techniques have been proposed. However, not all of them were effective enough. The constant development of these models and techniques show how these are important for accurate prediction. Our survey provides a comprehensive overview of some major software estimation approaches. These include parametric models, expert-based techniques, regression-based methods, analogy approaches, fuzzy logic, neural networks, and machine learning-based estimators. We reviewed research papers from 1970 to 2025 to get information of older models. But we mainly focused on recent papers from 2019 to 2025. We traced the progression and popularity trends of these models and discovered the increasing adoption of methods such as COCOMO, regression, and neural networks. Our study aims to help researchers and practitioners in understanding the landscape of software effort estimation models and techniques according to different generations with their pros and cons. It will also help in selecting or developing models ideal for some specific systems. Additionally, it highlights that no one model remains the best option for every software project.