A large language model (LLM) is a type of AI model trained on very large amounts of text data to generate, summarize, translate, and respond to language. LLM is a common synonym.
LLMs learn statistical patterns across billions of words. When given a prompt, they predict likely continuations based on those patterns. This produces fluent, coherent text — but the model has no factual memory, no access to real-time information (unless connected to external tools), and no understanding of truth in the way a human does.
An LLM is not a search engine — it does not retrieve indexed documents. It is not a database — it does not store structured records you can query. It generates text based on learned patterns.
Because LLMs predict plausible text rather than retrieving verified facts, they can produce confident-sounding statements that are incorrect or fabricated. This is called hallucination.