AI glossary · Models
GPT
In one sentence
GPT (generative pre-trained transformer) is OpenAI's family of language models. The name sums up how they work: they generate text, are pre-trained on huge amounts of data and are built on the transformer architecture.
Also known as: generative pre-trained transformer · GPT models · OpenAI GPT
What is GPT?
GPT is three words, and each one describes part of how it works:
- Generative: it produces new text rather than just classifying or searching.
- Pre-trained: before it’s used, it learns from huge amounts of text to predict the next word. It’s then tuned to follow instructions and hold a conversation.
- Transformer: the neural network architecture it’s built on, the transformer.
Put it to work
AI agents that look after your customers
From €49/mo + VAT
Voice, WhatsApp and email with the phone line included: agent, numbers and CRM in one fee.
See plans and pricingOpenAI released the first GPT in 2018 and has followed it with successive, more capable versions. The success of ChatGPT, launched in late 2022, led many people to use “GPT” as shorthand for AI in general. It isn’t: GPT is one family of LLMs among several.
Related to this term
UTIO agents
- Available 24/7 in 26 languages
- Set it up yourself for €0: built from your website
- 14-day free trial, no card needed
From €49/mo + VAT
See plans and pricingHow it works
A GPT model works like any generative language model:
- It takes your text and splits it into tokens.
- It works out, token by token, the most likely continuation.
- That choice includes an element of chance controlled by the temperature, which is why the same question can produce slightly different answers.
Recent versions in the family are multimodal: they can also interpret images or audio. Alongside them, OpenAI offers reasoning models that work through a problem step by step before answering. Names and versions change often, so check the provider’s documentation before you choose one.
Example
The owner of an estate agency in Brighton uses ChatGPT to write property listings. He feeds in the details (square footage, bedrooms, area, price) and gets an appealing description in seconds, which he then tweaks.
One day he asks for a listing for a flat “near the seafront” and the text says it’s “a five-minute stroll from the beach”. He never supplied that detail: the model filled it in because it sounded natural. That’s a hallucination. The lesson: a GPT model writes very well, but every figure and every promise to a buyer needs checking.
How it affects your business
If someone in your company says “we’ll do it with GPT”, it’s worth asking: which model exactly, through which application and with what data? The same task can often be done with models from other providers, or with smaller, cheaper models if it’s simple.
With UTIO you don’t have to pick a model to use an AI agent: the voice agent and the WhatsApp agent come with the language model built in and answer from your knowledge base, not from whatever the model remembers from the internet.
Frequently asked questions
Is GPT the same as ChatGPT?
No. GPT is the language model; ChatGPT is OpenAI's chat app that uses it. Think engine versus car: ChatGPT adds the interface, conversation history, web search and other tools on top of the GPT models.
Are all AI models GPT?
No. GPT is one specific family, from OpenAI. Other language models, such as Claude, Gemini, Llama or Mistral, also use the transformer architecture but come from other companies or communities.
What are GPTs?
That's OpenAI's name for custom versions of ChatGPT: a user adds instructions, documents and tools for a particular task. They still run on GPT models underneath.
Related terms
- LLM
An LLM (large language model) is an AI model trained on vast amounts of text to predict the next word. That single skill lets it draft, summarise, translate, classify and hold a conversation.
- Transformer
A transformer is a neural network architecture that processes every word in a text at once and uses an attention mechanism to work out which parts relate to each other. It's the foundation of almost every language model in use today.
- Language model
A language model is an AI system that calculates which word, or token, is most likely to come next in a piece of text. By repeating that prediction it can complete sentences, answer questions, summarise or translate.
- Chatbot
A chatbot is a program that holds text conversations with people, on a website or on WhatsApp, to answer questions or guide them through a task. It either follows fixed rules or uses a language model to understand what people write.
- Generative AI
Generative AI is artificial intelligence that creates new content, such as text, images, audio, video or code, from an instruction, using models trained on huge numbers of examples of that kind of content.
Updated on