Skip to content

Large Language Model (LLM)

A large language model is an AI model trained to process and generate language by predicting and working with sequences of tokens.

A large language model, or LLM, is an AI model trained to process and generate language by working with sequences of tokens.

LLMs can perform many language-related tasks through the same underlying model.

What an LLM Can Do

Depending on the model and system around it, an LLM can be used for:

The model does not need a separate hardcoded rule for every phrasing it encounters.

How It Works

At a simplified level:

Input
 ↓
Tokens
 ↓
Language model
 ↓
Predicted output tokens
 ↓
Response

The model has learned statistical relationships from its training process.

It uses the current Context Window and those learned relationships to generate the response.

LLM vs. Chatbot

An LLM is the underlying model.

A chatbot is an application or interface that may use an LLM.

A chatbot can also include:

The behaviour users experience therefore comes from more than the underlying model alone.

LLM vs. AI Agent

An LLM generates or interprets information.

An AI Agent combines a model with orchestration, tools, state, or actions so the system can complete multi-step work.

The model is a component of the agent rather than a synonym for it.

Why It Matters

Understanding that distinction helps when troubleshooting.

A poor answer might come from:

  • Model choice
  • Prompt
  • Missing context
  • Bad retrieval
  • Tool failure
  • Incorrect source data
  • Application logic

Simply changing the LLM does not necessarily solve the underlying problem.