Quick answer
A large language model is a neural network trained on huge amounts of text to generate and understand language. GPT-4, Claude and Llama are examples.
What it is
A large language model (LLM) is a neural network trained on huge amounts of text. It learns patterns in language and can write, answer, and reason. Modern LLMs also work with images, audio and code.
Why it matters
Most AI tools run on LLMs. Know what they do well and where they go wrong.
How to use it
- Pick a model based on the task: reasoning, summarisation, code, vision.
- Use system prompts to set behaviour and constraints.
- Ground the model in your own data with retrieval (RAG) when accuracy matters.
- Always have a human in the loop for high-stakes outputs.
Examples
- A chatbot uses an LLM to answer questions grounded in your product docs via RAG.
- A legal team uses an LLM to summarise contracts, with a lawyer reviewing every output.