Due to tremendous expansion in the banking, business, and IT sectors, data is growing at an exponential rate these days, and the majority of it is stored in relational databases. In order to access and alter information, users must be proficient with languages such as SQL (Structured Query Language). However, the customers are unfamiliar with SQL. As a result, a number of different models have been presented to date to convert English queries into SQL for extracting data from databases to achieve human performance on the dataset. It is an approach toward developing an end-to-end solution that may provide an intuitive text interface for many data sources within the corporate ecosystem, allowing them to obtain real-time data visualization or insights. Some of the models developed to date have already surpassed the 90 % test accuracy on the WikiSQL dataset. However, more research is needed to develop a largely scalable model with higher accuracy on custom and unknown databases. The purpose of this paper is to shed light on different strategies involved in this domain and to understand the contrast between them. We mainly focus on those factors and aspects that impact the performance and accuracy of the task significantly. This paper presents comprehensive research on the most prominent algorithms and analyses them based on different factors.