What is Large Language Model (LLM)?

An AI model trained on massive text data to understand and generate human-like language. LLMs like GPT, Claude and Gemini power most modern AI tools.

A large language model is a neural network trained on huge amounts of text, learning statistical patterns in language well enough to predict the next token in a sequence. That simple objective, scaled up with billions of parameters and massive datasets, produces models that can answer questions, write code, summarize documents, and translate. Most modern LLMs use the transformer architecture and go through two broad phases: pretraining on general text, then post-training (instruction tuning and feedback-based methods) to make them helpful and safe in conversation. LLMs do not look facts up in a database. They generate text from learned patterns, which is why they can be fluent and wrong at the same time. Families like GPT, Claude, Gemini, and Llama differ in size, training data, cost, and strengths, so the right choice depends on the task.

Example

When you ask ChatGPT to rewrite an email in a friendlier tone, an LLM reads your draft as tokens and generates a new version one token at a time.

Why it matters

Nearly every AI writing, coding, and chat tool is a wrapper around some LLM, so knowing which model powers a tool tells you a lot about its quality, cost, and limits. Browse the AI tools directory or the model leaderboard to put it into practice.