What is an LLM?

Understanding Large Language Models — the technology behind AI assistants.

A Large Language Model (LLM) is a type of generative AI trained on vast amounts of text. It learns patterns in language — how words and ideas connect — so it can predict what text should come next.

That is why LLMs are good at writing, summarising, translating, and answering questions. They do not truly "understand" like a human, but they are very skilled at producing plausible, useful language.

Examples include GPT (used by ChatGPT), Claude, Gemini, and Llama. When people say "AI assistant" or "AI chatbot" in a work context, they often mean an LLM-based tool.

LLM in action

Scenario: You paste a long contract clause into ChatGPT and ask for a plain-English summary.

The LLM reads your text, predicts useful language based on patterns it learned during training, and generates a summary. It does not look up the contract in a database — it produces language that fits your request.

Examples

•ChatGPT answering "Explain this contract clause in simple terms"
•An LLM summarising a 20-page report into bullet points
•Drafting a polite reply to a customer complaint
•Turning rough notes into a structured project brief

Key points

✓An LLM is AI trained on large amounts of text to predict language
✓LLMs power most popular AI chat assistants
✓They are skilled at language tasks but do not think like humans
✓GPT, Claude, and Gemini are all examples of LLMs

Knowledge check

What does LLM stand for?

Choose the best answer, then check your understanding.

Related dictionary terms

LLM