The cost perspective of adopting Large Language Model-as-a-Service

Large Language Models (LLMs) are pivotal in generative AI applications. Consequently, major cloud providers, such as Amazon, Azure, and Google, introduce the offering of LLM-as-a-Service (LLMaaS) products to enable businesses to leverage NLP, data analysis, and predictive modeling in their cloud solutions. This paper explores the incorporation of LLM-as-a-Service solutions into business workflows with a focus on inference costs. We review various LLMaaS offerings and conduct a comparative analysis based on a real-world case study of an AI chatbot.