Definition
What this term means
A type of artificial intelligence model trained on vast datasets of text to understand, generate, and reason about human language. LLMs power the AI assistants and generative search tools, including ChatGPT, Google Gemini, Claude, and Perplexity, that are rapidly becoming the primary way people discover products, services, and information online.
Why it matters
The business impact
LLMs are the engine behind the AI-driven shift in how people find and evaluate brands. Every recommendation an AI assistant makes is shaped by how the underlying LLM has learned to perceive your brand through its training data and real-time retrieval. Understanding how LLMs work, and what signals they rely on, is essential for any brand strategy that aims to remain visible in the AI era.
Used in context
How you might use this term
“A marketing team tested how five major LLMs described their brand category. They discovered that GPT-4 consistently recommended a competitor due to stronger training data signals, prompting a content strategy overhaul that improved their LLM representation across all platforms.”
Related terms
Explore connected concepts
GPT
OpenAI's conversational AI assistant, powered by the GPT family of language models. ChatGPT has become one of the most widely used AI platforms in the world, with over 200 million weekly active users. It handles everything from general knowledge queries and product recommendations to code generation and creative writing. ChatGPT can browse the web for current information and cite sources in its responses.
Transformer
A family of large language models developed by OpenAI that form the foundation of ChatGPT and numerous other AI applications. GPT models are pre-trained on internet-scale text data and then fine-tuned for conversational use, powering some of the most widely used AI assistants in the world. The architecture has become so influential that 'GPT' is often used colloquially to refer to generative AI in general.
Training Data
The massive datasets of text, code, and other content used to teach AI models during their initial training phase. Training data shapes the foundational knowledge of models like GPT, Gemini, and Claude, including what they know about brands, products, and industries. Sources include web crawls (such as Common Crawl), books, academic papers, Wikipedia, and publicly available databases.