This is a survey on deep learning models for text classification and will be updated frequently with testing and evaluation on different datasets. Natural Language Processing tasks ( part-of-speech tagging, chunking, named entity recognition, text classification, etc .) has gone through tremendous amount of research over decades. Text Classification has been the most competed NLP task in kaggle and other similar competitions. Count based models are being phased out with new deep learning models emerging almost every month. This project is an attempt to survey most of the neural based models for text classification task. Models selected, based on CNN and RNN, are explained with code (keras and tensorflow) and block diagrams. The models are evaluated on one of the kaggle competition medical dataset.