Operai announced Thursday that he has launched a previous view of research of his latest AI, GPT-4.5 model, for professional users and developers worldwide, with broader access that will be launched in the coming weeks.
GPT-4.5 comes with an improved ability to recognize patterns, generate creative ideas without reasoning and shows greater emotional intelligence, said the company.
Operai backed by Microsoft said it would begin to implement GPT-4.5 for Chatgpt Plus and team users next week, and then to Enterprise and Edu users the following week.
The CEO Sam Altman described him as “a giant and expensive model”, saying that the company was left without the capacity to deploy at user levels at the same time.
“We will add tens of thousands of GPU next week … This is how we want to operate, but it is difficult to perfectly predict the increase in growth that lead to GPU’s scarcity,” he said in an X publication.
Operai said the model also has a lower “hallucinations” rate, a common problem in which large language models generate false information. With 37.1%, the hallucination rate of GPT-4.5 is compared to the 61.8% shown by GPT-4O and 44% shown by its O1 reasoning model.
GPT-4.5 admits files of files and images and can also work in writing and coding projects, but currently does not admit other advanced features, such as voice and video.
