Short and sweet:
LLM, or large language model; the type of AI system, trained on massive amounts of text, that powers chatbots and answer engines like Claude, ChatGPT, Gemini, and Grok.
The underlying technology behind generative AI, rather than any single product or brand.
An LLM is trained by analyzing enormous volumes of text and learning the statistical patterns of how language is used, which lets it generate coherent responses, summarize information, and answer questions without being explicitly programmed with the answer to every possible query. Every major AI assistant, Claude, ChatGPT, Gemini, Copilot, Grok, is built on top of an LLM, though each company trains and fine-tunes its own model differently, which is part of why the same question can get a noticeably different answer depending on which tool is asked. For marketers, understanding LLMs at a basic level matters because it explains why answer engine optimization tactics work: a model is more likely to reproduce or cite information that appeared clearly and frequently across the text it was trained or grounded on.
A developer building a custom chatbot might choose which LLM to use as its foundation based on cost, speed, and accuracy tradeoffs.
A marketer explaining AI SEO to a client might describe an LLM in simple terms as the engine under the hood of tools like ChatGPT.
A content strategist noting that different LLMs cite their site inconsistently might adjust content structure to perform better across models generally, rather than optimizing for just one.
