OpenAI's GPT-4

Every previous AI model had clear limits — follow a complex instruction, and it'd miss half the details; ask for something creative, and you'd get something generic. GPT-4 is a different story.

OpenAI's GPT-4 is a large multimodal model that accepts both image and text inputs and produces text outputs. It excels at creative and technical writing tasks — composing songs, writing screenplays, learning your personal writing style — and handles nuanced, multi-part instructions with far greater accuracy than previous models. GPT-4 has demonstrated performance on a variety of human-designed benchmarks, including passing a simulated Uniform Bar Examination with a score in roughly the top 10% of test takers.

Available through the OpenAI API and ChatGPT Plus, GPT-4 represents a genuine leap forward in what AI can help you think through, create, and accomplish.