OpenAI’s Latest Version of ChatGPT, GPT-4o, is Free for Everyone

The free tier of OpenAI’s ChatGPT has just received a major upgrade, called GPT-4o (omni), launched at the company’s recent announcement event. As the name sort of implies, GPT-4o is an upgrade over GPT 4, complete with multimodal capabilities.

According to Mira Murati, OpenAI’s Chief Technology Officer, this updated model is significantly faster and enhances abilities in text, vision, and audio. Murati made these announcements during a live stream on Monday. The updated model will be free to all users, with paid users retaining up to five times the capacity limits compared to free users.

For now, the new GPT-4o has only received its text and image capabilities, but all of its features should be “rolled out iteratively”, as per a blog post from OpenAI.

Unlike GPT 3.5, GPT-4o is fully multimodal, meaning it can generate content and take input in the form of text, voice, and images, according to OpenAI CEO Sam Altman. Altman also mentioned on X that developers interested in experimenting with GPT-4o will have access to an API that is twice as fast and half the price of GPT-4 Turbo.

ChatGPT will be able to act as a “Her-like” (a 2013 movie about a conversational AI assistant) voice assistant once it gets new features as part of its voice mode. This will allow it to respond to users in real time and observe its surroundings through your camera. Currently, the voice mode has limitations, as it can only respond to one prompt at a time and relies solely on audio input.

The video below shows what it will be like. It shows two GPT-4os interacting with each other freely and even singing.