Which text classification model will suit a multi-class dataset with a large number of labels?

I have a dataset where there is a single sentence as input and a single label as output. There are over 40 different class labels with each label having a certain important keyword. For example: This phone is most durable in the market here durable is the keyword and it has a label X.

So far I have tried SVM but to no use, it fails in classifying well. What would be a good suggestion for a model to classify a multi-class dataset with a large number of labels

How many English words
do you know?
Test your English vocabulary size, and measure
how many words do you know
Online Test
Powered by Examplum