# What is GPT ?

GPT stands for **Generative Pre-trained Transformer**.

The following diagram explains GPT in a simple way:

![](https://cdn.hashnode.com/res/hashnode/image/upload/v1755024779395/1dea3e5a-8746-4bb2-a028-dde835c769ac.png align="center")

GPT (Generated Pre-Trained Transformer) Examples : GeminAI, ChatGPT etc

## Transformer :

Transformer is Just type of AI model architecture it uses.

The following diagram explains Transformer in a simple way:

![](https://cdn.hashnode.com/res/hashnode/image/upload/v1755025168017/155a0ffd-9be1-4767-9d79-c4e3fa1a82de.png align="center")

also, the Transformer architecture was indeed first introduced by Google researchers **<mark>in the 2017 paper "Attention is All You Need"</mark>**

refer this link : [https://papers.neurips.cc/paper/7181-attention-is-all-you-need.pdf](https://papers.neurips.cc/paper/7181-attention-is-all-you-need.pdf)

## Conclusion :

In general, GPT is just a Generative Pre-trained Transformer. It works on a simple concept: the user gives input, the transformer processes it into tokens, and then provides the output — nothing more.