OpenAI releases GPT-4 with multimodal input

Model release OpenAI Source language: English

OpenAI introduced GPT-4, highlighting stronger results on professional exams and complex tasks, and brought image input into its flagship model for the first time. Model evaluation began to focus on reasoning and reliability rather than parameter count alone.

Excerpt

OpenAI introduced GPT-4, highlighting stronger results on professional exams and complex tasks, and brought image input into its flagship model for the first time. Model evaluation began to focus on reasoning and reliability rather than parameter count alone.

Short excerpt from the source — open the original for the full text.

Sources

← Back to the daily brief