📚 Glossary
GPT
In one line: Generative Pre-trained Transformer. The model architecture and naming convention used by OpenAI for ChatGPT.
GPT stands for Generative Pre-trained Transformer. It's both a model architecture and a naming convention used by OpenAI:
- Generative — produces new text rather than classifying
- Pre-trained — trained on a massive dataset before being used
- Transformer — uses the transformer architecture (attention-based)
GPT versions: GPT-1 (2018), GPT-2 (2019), GPT-3 (2020), GPT-4 (2023), GPT-4o (2024), GPT-4.1 (2025), GPT-5 (2025+).
While 'GPT' is OpenAI's branding, the underlying transformer architecture is used by every modern LLM — Claude, Gemini, DeepSeek and Llama all use transformers.
See it in action — ask any AI about gpt on AskAI.free.
Try it free →