On Tuesday, OpenAI broadened access to its image-generation feature, now powered by the GPT-4o model, to all users. This enhancement follows the initial rollout to paid subscribers of ChatGPT last month, allowing even those on the free tier to participate. Free users, however, will face some restrictions, including a lower usage cap. OpenAI, headquartered in San Francisco, emphasized that images generated via this feature will include metadata containing Coalition for Content Provenance and Authenticity (C2PA) information to help counter deepfake incidents.
OpenAI Expands GPT-4o Image Generation to Free Users
In a recent post on X, previously known as Twitter, OpenAI CEO Sam Altman announced the availability of the GPT-4o-powered image-generation tool for free-tier users of ChatGPT. Users must be logged into their OpenAI account to access the feature, which currently limits free users to generating three images per day, along with a cooldown period between consecutive image creations.
In another post, Altman shared the remarkable achievement of one million new registrations on ChatGPT within a single hour on Monday. This impressive figure surpassed the platform’s earlier record of adding one million users over the course of five days at its launch in 2022.
The surge in interest for the image-generation feature is largely attributed to a viral trend where users are employing it to transform their photos into the art styles associated with Studio Ghibli, the renowned Japanese animation studio behind classics like Spirited Away and Princess Mononoke. Social media feeds have been inundated with users sharing their experiences in this “Ghibli style.”
This trend’s popularity prompted Altman to impose temporary rate limits on the paid version of ChatGPT when generating images due to the overwhelming demand, humorously stating that “our GPUs are melting.”
In a different context, Altman recently criticized Meta for its restrictions on the use of its open-source Llama models by companies with over 700 million monthly users. In a post, he expressed, “We will not do anything silly like saying that you can’t use our open model if your service has more than 700 million monthly active users.”
The new image generation capabilities in ChatGPT also feature enhanced text rendering and a prompt-based inline editing tool. Users can submit their own images and request transformations into various artistic styles or make edits. OpenAI has implemented measures to block any requests for images containing harmful content, including child sexual abuse material, sexual deepfakes, or any violent imagery.
