This research paper investigates the diversity in AI-generated responses to identical user queries, with a focus on lexical diversity. The study evaluates various AI models, including OpenAI’s ChatGPT, Google’s Bard, and Gemini, by measuring variations in vocabulary richness using the Type-Token Ratio (TTR). By examining response diversity, this study aims to assess and analyse AI-generated responses using the Type-Token Ratio (TTR), this study aims to compare the linguistic diversity across three leading AI models: ChatGPT, Bard, and Gemini and how AI models maintain creativity, coherence, and adaptability in user interactions. Additionally, graphical representations of data findings provide insight into diversity trends across AI platforms.