Automated short answer grading (ASAG) is a crucial task in educational technology, aimed at providing timely and consistent evaluations of students’ responses. This paper presents an ASAG system utilizing Fasttext word embeddings to enhance the accuracy and efficiency of grading short answers. This paper presents a state-of-the-art approach for automated short answer grading (ASAG) using word representation in Fasttext. Compute the learner grade based on threshold value and compare the vector representation of each student's answer to a model answer using similarity metrics. Also explained the accuracy of the model with precision, recall, F1-score and highlighting how well it captures the entire context of the answers.