Glossary Concept
Large Language Model (LLM)
A large language model is a machine learning model trained on large text and code datasets to predict and generate text, code, summaries, and other structured output.
Large Language Model (LLM) is the broad term for models such as GPT, Claude, Llama, Mistral, and similar systems that generate text or code from a prompt.
An LLM does not understand a system the way a person does. It predicts likely output from training and context. That can be incredibly useful for drafting, summarising, explaining, and coding, but it can also produce confident nonsense.
The practical rule is simple: use the model for leverage, not authority. Give it good context, check the output, and be especially careful when it touches security, money, production systems, or anything you cannot easily undo.