OpenAI Releases GPT-5
Two and a half years after GPT-4, OpenAI's new flagship folds its fast chatbot and its slow reasoning model into one system that decides how hard to think. The reception was more muted than for past releases — a sign, some argued, that progress is becoming steady rather than shocking.
OpenAI released GPT-5 today, the successor to the model that turned ChatGPT into a phenomenon two and a half years ago. Like Anthropic's recent releases, it is a single system that combines quick responses with slower step-by-step reasoning, deciding on its own how much thought a question deserves — so users no longer choose between different models. It is better than GPT-4 across the board, OpenAI says, and notably better at coding, at writing, and at not making things up, and it is available to all ChatGPT users, including the free ones, in some form.
The interesting thing about the release was the reaction, which was calmer than any major OpenAI launch since 2022. GPT-4, in 2023, felt like a jump into new territory. GPT-5 felt, to many who tried it, like a very good version of what already existed. Some took that as disappointment — evidence that simply scaling these models up is yielding less than it used to, and that the field may be approaching a plateau. Others took it as maturity: the technology is becoming a steadily improving tool, the way each year's phone is better than the last without being a revelation.
Sam Altman had spent months managing expectations in both directions, at times comparing the new model to a jump and at times warning against hype. The muted response is itself a kind of data point. When ChatGPT launched, any improvement was astonishing because the baseline was zero. Now the baseline is high, and a better model has to clear a bar that the technology itself keeps raising.
Whether the calm reaction reflects a real slowdown or just a jaded audience is the question underneath, and it is the same question the whole industry is now being asked by investors: is this still accelerating, or leveling off?