Definition
An LLM (large language model) is an AI model trained on huge amounts of text to generate language, such as the models behind ChatGPT, Claude and Gemini.
LLM explained
A large language model learns patterns from a huge body of text during training, then generates a response piece by piece, based on what is most likely to come next given the prompt. The model itself has no live connection to the web: what it knows comes from its training data, which stops at a cut-off date.
That is why most AI assistants now combine the model with retrieval. When you ask ChatGPT with search, Perplexity or Google's AI Mode a question, the system searches the web or an index, reads a handful of pages, and asks the model to write an answer grounded in those pages, often with citations.
For marketers this creates two routes into AI answers. The slow route is being well represented in the text models are trained on: consistent facts about your brand across many reputable sources. The fast route is being retrieved at answer time: crawlable, well-structured pages that rank for the searches the assistant runs behind the scenes. A good GEO plan works on both.
Example
Ask an assistant with web search turned off what your company does, and it answers from training data, possibly using an old product name. Ask again with search turned on, and it reads your current About page. Your plan for AI visibility needs to cover both situations.
Why it matters
Every AI search product runs on an LLM. Knowing how models learn and retrieve information explains why brand consistency and crawlable pages both matter for AI visibility.
Related service
Generative Engine Optimization (GEO)
We help your brand get found, understood and cited by ChatGPT, Perplexity, Gemini and Google AI Overviews.
Generative engine optimization servicesPublished by Vidern, founded and led by Malhar Shah. Updated .
See how AI assistants describe your brand
The free GEO audit scores your readiness for ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews from 0 to 100, with an action plan, within 24 hours.
Related terms
- RAG (Retrieval-Augmented Generation)RAG (retrieval-augmented generation) is when an AI system first retrieves relevant documents, then has a language model write its answer from them.
- AI hallucinationAn AI hallucination is a confident but false statement from a language model, such as an invented fact, quote, citation or product detail.
- LLM citationAn LLM citation is a link or named source that an AI assistant shows alongside its answer to indicate where the information came from.
- GEO (Generative Engine Optimization)GEO (generative engine optimization) is making your brand easy for AI assistants such as ChatGPT, Perplexity and Gemini to find, trust and cite.
- Context windowA context window is the most text, measured in tokens, that a language model can consider at once, including instructions, documents and its reply.