Large Language Models (LLMs) rely heavily on prompts as a means of communication and guidance. Users can use them as inputs or queries to a model to get it to produce desired results.