Large Language Models (LLMs) are computational models known for their ability to achieve general-purpose language understanding and generate text by training the model’s billion parameters on massive amounts of text data. Multilingual LLMs are language models that support multiple languages and can understand and generate text in different languages. In the current scenario, there is a lack of Multilingual LLMs that have dedicated support for Indian languages. This paper has tried to assess the current performance of LLMs in Indian languages by hand-picking a few multilingual LLMs (like XLM-R, mBERT, MuRIL, Navarasa, IndicBERT v2, mT5) that support these Indian languages and fine-tuning them using the technique of prompt tuning and few-shot learning for multilingual Sentiment Analysis. Sentiment Analysis involves processing text to determine whether the emotional tone of the message is positive, negative, or neutral. It requires a model to understand and interpret the context, tone, and implicit meaning of text, which provides a robust test of the model’s natural language understanding capabilities.