OpenAI released a new AI image generator last week, the GPT-4o model that works directly inside ChatGPT to create stunning images. The tool became an overnight sensation. In a matter of hours, the web was flooded with AI-generated content from the 4o model, including a massive amount of Studio Ghibli-style creations and deepfakes featuring celebrities.
OpenAI has been enjoying all the free publicity, with the 4o image generator tool making ChatGPT viral at a time when DeepSeek and Google each launched their newest AI models. OpenAI has also largely ignored valid criticism concerning the ease of creating potential copyright-infringing images and deepfakes that can mislead people. Instead, OpenAI doubled down on the new image generator in ChatGPT, saying it prefers to allow the model to be as free as possible rather than impose stricter safety features.
The downside of ChatGPT’s new image generator becoming so widely used was that ChatGPT Free users did not get access to the new tool as initially expected. Users on the free tier were supposed to be able to create AI images with the GPT-4o model at the same time as ChatGPT Plus, Pro, and Team users.
OpenAI Sam Altman said on X last week that due to the massive demand for the 4o image generator, ChatGPT Free users would have to wait — but the wait was much shorter than expected. I didn’t think the company to be ready to bring its new mage generator to ChatGPT Free users so soon, but it just happened.
Alman took to X on Monday to give the world an idea of how popular the new ChatGPT feature is. “The ChatGPT launch 26 months ago was one of the craziest viral moments I’d ever seen, and we added one million users in five days,” he said. “We added one million users in the last hour.”
That remark proves once again how incredibly popular the new image generator is. There’s nothing else in ChatGPT’s arsenal of AI tools that would make the service so appealing to users. Image generation must be the reason why so many people flocked to create new accounts.
I’m assuming here that Altman is talking about ChatGPT account holders, of course. It’s unclear how many of those are paying subscribers, but that doesn’t even matter as long as OpenAI is onboarding new users.
After sharing additional Project Ghibli-style ChatGPT-generated images centered around India on X, the CEO announced that ChatGPT image generation is rolling out to all Free users.
The reveal dropped last evening, though Altman did not mention additional details that might matter to Free users.
A few days ago, he said that some limitations might impact Free users, like the ability to make only three images with AI per day. At the time, Altman said that OpenAI’s GPUs were “melting” in response to the demand from ChatGPT users to generate AI images with the new AI tool.
With OpenAI raising another $40 billion at a $300 billion valuation, of which $10 billion is contingent upon the company becoming a for-profit entity by the end of the year, it makes sense for Altman & Co. to want to attract new users to ChatGPT.
Giving them free access to the 4o image generation tool is a great way to do this. Some people might come for the Studio Ghibli memes and celebrity deepfakes but might end up staying for other complex AI features and even paying for access.
Meanwhile, ChatGPT users who pay a subscription reacted to Altman’s announcement. Some complained they’re seeing limits on their paid tiers, so Free users shouldn’t get it. Others complained that OpenAI apparently added more safety features to the ChatGPT image generator, thus making it less likely to generate imagery as freely as in the first days.
Some ChatGPT Plus users found that ChatGPT lets them generate about five images per hour before locking access. That’s still better than what’s available for ChatGPT Free users.
Then again, the $40 billion the company just raised should help OpenAI improve its infrastructure and potentially acquire more GPUs to sustain the 4o image generation frenzy.
