What Is an LLM? Large Language Models Explained
A large language model (LLM) is software trained on vast amounts of text to predict what comes next in a piece of writing. That simple skill, at huge scale, lets it answer questions, write, summarise and code.
What is it?
“Large” refers to the size of the model and the amount of text it learned from. “Language model” means it models language: given some words, it works out which words are likely to follow. ChatGPT, Claude and Grok are apps built on top of LLMs.
How does it work, roughly?
- Training. The model reads a huge collection of text and adjusts itself millions of times to get better at predicting the next piece.
- Fine-tuning. It’s then taught to be helpful, follow instructions and avoid harmful output, often with human feedback.
- Using it. When you type a prompt, it produces a reply one token at a time, each chosen based on everything so far.
The model is built to produce plausible text, not to look facts up in a database. That’s why it can sound sure while being wrong. This is called a hallucination.
Models vs products
A model is the trained engine (for example “Claude Sonnet 5.5” or “Grok 4.7”). A product is the app you use around it (Claude, ChatGPT, Grok), which adds features such as file upload, memory, web search and projects. Companies release new models often, so a product’s model changes over time. We track the latest on the homepage AI Status card.
Common mistakes
- Thinking it “knows” today’s news. Unless it has web search switched on, it only knows what it learned in training.
- Assuming a bigger model is always better. Smaller models can be faster and cheaper, and good enough for simple jobs.
What to try next
What is a context window? explains how much an LLM can hold in mind at once.
Sources
- General explainer written by Silent Overseer; concepts are standard and not time-sensitive.
Silent Overseer uses AI tools to help research and draft, and every article is reviewed before publishing. How we work. Spotted a mistake? We correct and date changes.