What is Large Language Model (LLM)?
A large language model (LLM) is an AI system trained on large volumes of text to generate human-like language, and is the underlying technology behind AI chat and search products like ChatGPT, Gemini, and Perplexity.
An LLM generates responses by predicting likely sequences of text based on patterns learned during training, rather than by looking up facts in a fixed database — which is part of why techniques like retrieval and citation (grounding a response in specific, checkable sources) matter for accuracy and trust in AI search products.
Different products layer different retrieval and search capabilities on top of an underlying LLM, which is why the same page can be extracted, cited, or ignored differently across ChatGPT, Gemini, and Perplexity even when asked a nearly identical question — each system's retrieval layer and training data differ.