Automatic natural language processing is a dynamic study domain to which numerous scholars have contributed, persistently striving to enhance result accuracy. One of the uses is the automated creation of multiple-choice questions. This article introduces a system for the automated creation of multiple-choice questionnaires across numerous steps to resolve this issue. Initially, we examine the source text utilizing keyphrase extraction techniques. Subsequently, we implement a fine-tuning procedure on a large language model, including Text-to-Text Transfer Transformer (T5), Bidirectional and Auto-Regressive Transformer (BART), and Generative Pre-trained Transformer (GPT), to provide pertinent questions derived from these key phrases. Subsequently, we employ techniques to create distractors. This facilitates the automatic generation of multiple-choice questionnaires, minimizing the time and effort required for assessment preparation. The code was run in an environment equipped with 29 GB of RAM and a P100 GPU.