A large multimodal language model released by OpenAI in 2023, capable of accepting text and image inputs and producing text outputs, and a successor to the GPT-3.5 series.
Semantic Classification
Content
- GPT-4 is a transformer-based language model trained on large text corpora and aligned with human feedback. It accepts longer context windows than earlier GPT releases and can process images alongside text in its multimodal configuration.
- OpenAI did not publish full architectural or training details, so public knowledge rests on the technical report and observed behaviour. The model is offered through an API and consumer products, and is used for tasks including drafting, summarisation, coding assistance and question answering.