Large language model (LLM)
An AI model trained on huge amounts of text so it can read and write language. It works by predicting which words are likely to come next.
A large language model is the engine behind most of today's AI chat tools. During training it reads an enormous amount of text, from books and websites to code, and learns how words tend to follow one another. "Large" refers both to that amount of text and to the size of the model itself, which has billions of adjustable settings.
When you ask an LLM to write a polite email declining an invitation, it builds the reply one small piece at a time, each time picking a likely next word. The result often reads as if a person wrote it.
Fluent writing is not the same as knowing the facts. An LLM can produce a smooth, confident answer that is wrong, which is called a hallucination. Treat its answers as a helpful first draft, and check anything that matters.
Related: Model (AI model), Chatbot, Token, Training data