Large Language Model (LLM)

What is a large language model?

A large language model (LLM) is an AI model trained on very large amounts of text to understand and generate human language. Well-known examples are OpenAI’s GPT models behind ChatGPT, Google’s Gemini, Anthropic’s Claude and Meta’s Llama. LLMs power chat assistants, AI search features such as AI Overviews and AI Mode, and many marketing tools.

Two ways in which LLMs “know” things matter for marketers:

• Training data: during training, a model learns from huge amounts of web pages, books and other text up to a cut-off date. What it knows about your brand then depends on how often and how consistently it was mentioned (brand mentions).
• Retrieval: AI search tools also look up current web pages at the moment of the question (retrieval-augmented generation) – often by running several searches at once (query fan-out) – and cite them as sources (AI citations).

LLMs work in many languages, but their quality and the sources they rely on differ by language and market. An answer about your industry in German or Spanish may draw on completely different websites than the same question in English.

In marketing, LLMs are used for research, drafting, localization and data analysis. Their output still needs expert review: they can state wrong facts convincingly (“hallucinations”) and miss local nuances.