GPT-4o

GPT-4o ("omni") is a single model that accepts any combination of text, audio, image, and video inputs and generates text, audio, and image outputs. It responds to audio inputs in as little as 232 milliseconds, averaging 320 milliseconds, which is similar to human conversational response time. The article says it matches GPT-4 Turbo on English text and code, improves on non-English text, and is 50% cheaper in the API. Text and image input with text output is rolling out now, with audio capabilities to follow in coming weeks and months.