In May 2024, OpenAI announced GPT-4 Omni, a multimodal language model that could "accept as input any combination of text, audio, image, and video and generates any combination of text, audio, and image outputs."

Will OpenAI make available to the general public or to Tier 5 developers any "GPT-4" model that is capable of taking as input in the same datastream text, audio, and image or video and outputting text, audio, and images?

Currently, advanced voice mode does not support image or video input and cannot output images.