OpenAI expanded its image-generation feature powered by the GPT-4o model to all users on Tuesday. The artificial intelligence (AI) image generation capability was first rolled out to the paid subscribers of ChatGPT last month, and now even those on the free tier can access it. However, the free users will get limited usage with a high rate limit added. The San Francisco-based AI firm stated that each image generated with this feature will include Coalition for Content Provenance and Authenticity (C2PA) information into the metadata to mitigate instances of deepfakes.
OpenAI Expands GPT-4o Image Generation to Free Users
In a post on X (formerly known as Twitter), OpenAI CEO Sam Altman announced that the GPT-4o-powered image-generation feature in ChatGPT is now available to those on the free tier of the platform. Notably, users should be logged into their OpenAI account to use the feature. Currently, the free tier has a rate limit of three image generations per day. Additionally, there is a cooldown period between generating two images consecutively.
In a separate post, Altman highlighted that one million new users registered on ChatGPT in just one hour on Monday. He also compared it to the platform’s previous record of adding one million users in five days when it first launched in 2022.
One of the main reasons behind the popularity of the new feature is the viral Internet trend where users are using the ChatGPT feature to turn their images into the art style of Studio Ghibli, the popular Japanese animation studio behind anime movies such as Spirited Away, Princess Mononoke, and My Neighbour Totoro. Social media platforms have been flooded with people posting their memorable moments in the “Ghibli style”.
The popularity of this trend was so strong that Altman added temporary rate limits for the paid ChatGPT tiers when generating images using the AI chatbot. The OpenAI CEO said the decision was taken as “our GPUs are melting.”
Separately, Altman recently took shots at Meta, which now restricts the usage of its open-source Llama models by companies that have more than 700 million monthly active users. He said in a post, “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.”
Notably, the new image generation in ChatGPT comes with improved text rendering and a prompt-based inline editing feature. Users can also add their images and ask AI to transform them into various styles or to make edits. OpenAI said the model blocks requests for images that include harmful content such as child sexual abuse material, sexual deepfakes, or violence.
