A large language model (LLM) is an AI model trained on vast amounts of text to predict and generate language. It is the core technology behind AI assistants like ChatGPT, Claude, and Gemini, producing answers by predicting the most likely next words given a prompt and its training.
For brands, the key implication is that an LLM's answers reflect two things: what it absorbed during training, where consistent brand presence across the web matters, and what it retrieves live at query time. Understanding both pathways is the foundation of improving AI visibility.