OpenAI unveils its new flagship, ChatGPT-4o

OpenAI’s latest ChatGPT flagship is ChatGPT-4o (“o” for “omnimodel”) which will operate across images and audio as well as text, and will be available to all users, including free users. The news came in a brief online announcement from Mira Murati, OpenAI’s CTO and former interim CEO during Sam Altman’s brief removal from the role. Altman did not appear in the online session.

The announcement was followed by demos, especially representing ChatGPT-4o’s conversational abilities. The entire session lasted barely 30 minutes, and despite rumors there was no discussion of ChatGPT for search.

Chatting up a storm. In addition to showing off its mathematical abilities and skill at analyzing complex charts, the new model proved to be an entertaining and versatile conversationalist. Users can now interrupt the audio in mid-flow and the model will respond immediately.

The model was invited to tell a bedtime story on a specific subject, was constantly interrupted with suggestions for voice and tone, and proved highly responsive.

Why we care. For marketers, the new model provides an enhanced version of previously existing capabilities. Murati emphasized an improved UI for greater ease of use. She said the new flagship is “much faster” and also said that it now has a memory for previous queries and prompts from each user, creating some continuity.

Free users will now be able to take full advantage of visual and audio capabilities, but paid users will continue to have preferred access to the platform. Murati’s announcement that an API will be available to developers was greeted with applause.