OpenAI’s new picture generator, powered by its GPT-4o mannequin, is now out there to all customers, CEO Sam Altman stated in a submit on X. The function was till now out there solely to paying customers of ChatGPT.
Whereas it isn’t clear what number of photographs customers on the free tier can generate, Altman final week had talked about a restrict of three photographs per day.
OpenAI’s picture era instrument took off immediately after launch, with Altman saying the demand was so excessive, the corporate’s GPUs had been “melting.” The instrument rapidly additionally gained notoriety for getting used to transform photos into the fashion of Japanese animation agency Studio Ghibli, elevating considerations round copyright and coaching knowledge utilized by the corporate as properly, given the similarity in fashion.
Some individuals additionally used it to generate pretend receipts, corresponding to restaurant payments. An OpenAI spokesperson instructed TechCrunch that each one these photographs have metadata indicating that ChatGPT generated them, and that the corporate “takes actions” if the photographs violate the corporate’s tips.
In the meantime, OpenAI as we speak stated it raised $40 billion in funding led by SoftBank at a $300 billion valuation. The corporate additionally stated ChatGPT has hit 500 million weekly energetic customers and 700 million month-to-month energetic customers.