What Is an LLM?
A Non-Technical Guide

You've heard of ChatGPT, Claude, and Gemini. They're all powered by something called a "Large Language Model" or LLM. Here's what that means in plain English.

The Simple Explanation

An LLM is a computer program that has read a huge amount of text — books, websites, code, conversations — and learned the patterns of language from it. When you ask it a question, it predicts what a helpful response should look like, word by word.

Think of it like a very sophisticated autocomplete. Your phone's keyboard predicts the next word you'll type. An LLM does the same thing, but trained on trillions of words, making its predictions remarkably good.

What "Large" Means

The "Large" in LLM refers to two things:

• Training data: Trillions of words from books, websites, code repositories, and more.
• Model size: Billions or trillions of "parameters" — adjustable numbers that store what the model has learned. GPT-4 is estimated to have over 1 trillion parameters.

More parameters generally means the model can capture more nuanced patterns, but it also means it costs more to run.

What LLMs Can Do

• Write and edit text: Emails, articles, reports, creative writing.
• Answer questions: From trivia to complex technical questions.
• Write code: Create, debug, and explain programming code.
• Analyse data: Summarise documents, extract information, compare options.
• Translate languages: Between dozens of languages with high accuracy.
• Reason through problems: Newer models can break down complex problems step by step.

What LLMs Cannot Do

• Know facts with certainty: They can "hallucinate" — generate plausible but incorrect information.
• Access the internet in real time: Unless specifically given that ability, they only know what they were trained on.
• Feel emotions or have opinions: Despite sounding natural, they don't have feelings or consciousness.
• Replace human judgement: They're tools to help you think, not to think for you.

The Main LLMs in 2026