Openai has introduced a new feature that gives Chatgpt users more control over how fast GPT-5 reacts with thinking. The update, which was announced at X on September 18, enables Plus, Pro and Business subscribers to select the ‘thinking time’ of the model directly from the message composer on the web. Openai adds previously speed control to GPT-5 thinking. However, many users have said they want more flexibility to balance depth and speed, depending on the task. Among the new system, subscribers can now vary between different fitting modes: Standard: The new standard for all plans, which offers a balance between speed and intelligence. Extensive: The former standard for Plus users, where the model takes more time to provide deeper answers. Light: Exclusive to Pro users, providing the most delicious answers for quick tasks. Heavy: Also exclusive to Pro, designed for the deepest reasoning, with slower but more comprehensive outputs. Openai says the selected mode will remain active on future chats on the web until users change it by hand. “We’ve heard your feedback that GPT-5 thinking can sometimes take longer than you want. Now plus, pro and business users can set the pace to fit the moment,” the AI company wrote on its X account. GPT-5 Codex for developers, meanwhile, also unveiled Earlier this week GPT-5 Codex, a new version of its AI specifically designed for software development. The model is built to help developers write, debug and revise code more efficiently, and work interactively with people and independently of complex tasks. Openai claims that GPT -5 Codex can handle a wide range of programming tasks, from creating projects from scratch to adding features, correction of errors and the control of other people’s code. It can work on long projects for hours, and test and improve its own code until it works properly. For smaller, simpler tasks, it responds quickly without wasting resources. The AI is trained to review code carefully and find critical errors before causing problems, Openai said in his blog post. Tests on popular Open Source projects have shown that the proposals are more accurate and useful than before, which help developers focus on the main problems.
Openai’s latest update makes Chatgpt users choose how fast GPT-5 thinks: How the new feature works
