Large Language Model (LLM)

A large language model is a generative model trained on very large amounts of text to predict and produce language.

In detail

An LLM has no live view of the web. It knows what was in its training data, plus whatever is supplied to it at the moment of the request — search results, a pasted document, a retrieved page.

That is why retrieval matters: for anything current or specific, the model is only as good as the sources it can fetch and read.

Example

A chat assistant answering a question about a product page is usually reading a fetched version of that page, not remembering it.

Why it matters

Content that is blocked, JavaScript-dependent or vague is hard to retrieve and quote, whatever its quality.

Common mistakes

Optimising for a model's "memory" instead of for retrievability.
Assuming a model can see pages it has no access to.

Related guides

The Complete AI SEO Guide
Guide
How to stay visible when answers are generated: retrievability, answer-shaped content, entities, structured data, and honest measurement.

Related terms

Generative AI
Generative AI is a model that produces new content — text, images, code or audio — rather than only classifying existing data.
Retrieval-Augmented Generation (RAG)
RAG is the pattern of retrieving relevant documents first and then having a language model answer using them.
AI Search
AI search is any search experience where a generative model answers the query directly instead of only listing links.
Prompt Engineering
Prompt engineering is the practice of writing instructions that get reliable, useful output from a generative model.