GPT-4o, OpenAI's new multimodal artificial intelligence, promises faster and enhanced capabilities.

OpenAI has just unveiled GPT-4o, an improved version of its GPT-4 model, which powers the company's flagship product, ChatGPT. During a live announcement on Monday, OpenAI's CTO, Mira Murati, stated that this updated model is 'much faster' and improves 'text, vision, and audio capabilities'.

GPT-4o will be free for all users, while paying users will continue to benefit from 'capacity limits up to five times higher' than those of free users, Murati added.

Gradual Deployment of Capabilities

In a blog post, OpenAI specifies that GPT-4o's capabilities 'will be deployed gradually,' but its text and image features will start to become available in ChatGPT from today.

A Multimodal Model

OpenAI's CEO, Sam Altman, posted that the model is 'natively multimodal,' meaning it can generate content or understand commands in voice, text, or images.

Developers who want to experiment with GPT-4o will have access to the API, which is offered at half the price and twice as fast as GPT-4 Turbo, Altman added on X.

New Features in Voice Mode

New features are coming to ChatGPT's voice mode thanks to this new model. The application will be able to act as a voice assistant like 'Her,' responding in real-time and observing the environment around you.

The current voice mode is more limited, responding to a single command at a time and only processing what it can hear.

Reflections on OpenAI's Trajectory

Altman reflected on OpenAI's evolution in a blog post after the live event. He stated that the company's initial vision was to 'create all kinds of benefits for the world,' but acknowledged that this vision had changed.

OpenAI has been criticized for not open-sourcing its advanced AI models, and Altman seems to indicate that the company's goal has shifted to providing these models to developers via paid APIs, leaving it to these third parties to create applications.

'It now seems that we will create AI, and others will use it to create all kinds of amazing things from which we will all benefit.'

Speculations and Announcements

Before the launch of GPT-4o, various reports predicted that OpenAI would announce an AI search engine to rival Google and Perplexity, a voice assistant integrated into GPT-4, or an entirely new improved model, GPT-5.

Naturally, OpenAI carefully scheduled this launch just before Google I/O, the tech giant's flagship conference, where various AI products from the Gemini team are expected to be launched.