OpenAI has made its newest major language model, GPT-5, available to all ChatGPT users and developers. The company’s CEO, Sam Altman, stated that GPT-5 is significantly more advanced than previous models. Altman likened it to the first iPhone with a Retina display, describing it as an irreversible improvement.

GPT-5 Launched

The new model is reported to be faster, more accurate, and delivers significantly fewer errors than its predecessor, GPT-4. Altman stated that speaking with GPT-3 is like chatting with a high school student, while GPT-4 is university-level, and GPT-5 feels like speaking with a doctoral-level expert for the first time. According to Altman, GPT-5 delivers top performance not only on general knowledge questions but also in many other areas, such as coding, writing, and healthcare.

GPT-5 is available as a single model on the ChatGPT platform. Instead of the previous version, which offered users different models and “advanced reasoning” options, GPT-5 operates with an automated prompter running in the background.

This prompter switches to different logic modes when faced with complex questions or when the user issues commands like “think.” Altman described the previous interface as “cluttered and confusing.”

During the launch of GPT-5, it was also announced that ChatGPT had reached 700 million weekly users. Despite this, the company had recently lost leadership in the field of artificial intelligence to other players. With GPT-5, OpenAI aims to regain this leadership. The company claims that this model is superior not only to text writing or code generation, but also to high-precision fields like healthcare.

Nick Turley, head of ChatGPT, said that the new model “feels very good.” It is stated that OpenAI prioritized user experience, not just performance and accuracy, in the development of this model.

Some details emerged on GitHub early in the day before the launch of GPT-5. However, the official announcement was made by OpenAI on August 7, 2025. The company also simplified the user interface to eliminate the confusion experienced with previous models.

With the new model, users are expected to experience a more seamless, natural, and expert-level AI experience. Initial feedback and performance analyses regarding the use of GPT-5 are expected to be released in the coming days.