What Is an LLM?

What Is a Large Language Model?

You may have heard people talk about tools like ChatGPT, Claude, or Gemini. Some people use them to write emails, answer questions, explain homework, or help with computer code.

Behind these tools is a type of technology called a Large Language Model, or LLM.

The name sounds technical, but the main idea is actually simple.

An LLM is a computer system that has read a very large amount of text. From that text it learns how language is usually written and how sentences normally continue.

When you ask it a question, it writes a response by continuing the text step by step.

---

The Core Idea

Imagine you start a sentence like this:

The capital of France is

Most people would continue that sentence with the word Paris.

Why?

Because we have seen that sentence many times before in books, school lessons, or websites.

A Large Language Model works in a similar way. During training it reads enormous amounts of writing—books, articles, websites, conversations, and documentation.

From this training it learns patterns in how sentences continue.

When you type a prompt, the model does not search for a stored answer. Instead, it looks at the text you gave it and predicts what piece of language is most likely to come next.

It writes that piece of text.

Then it looks again and predicts the next one.

And then the next.

Little by little, those small predictions turn into a full response.

---

A Simple Example

Imagine typing this message:

I will call you when I

Many people would continue the sentence with something like:

• arrive
• get home
• finish work

Your brain is predicting how the sentence will probably continue.

A Large Language Model does the same thing, but it has learned from a much larger amount of text. Because of that training, it can continue writing for many sentences or even several paragraphs.

---

The One Idea That Explains LLMs

Everything an LLM does comes from a very simple loop:

predict the next piece of text
add it to the sentence
repeat

The model keeps repeating this process until the answer is complete.

Because it has seen so much writing during training, the responses often sound natural and well explained.

But the system is still doing the same basic task each time: predicting what text should come next.

---

Summary

A Large Language Model (LLM) is a computer system trained on massive amounts of text. It generates answers by predicting what piece of language should come next and repeating that process many times.

Even though the idea is simple, the large amount of training data allows these models to produce responses that sound fluent and helpful.

In the next lesson, we will look at how computers break language into small pieces so that models can work with it.