OpenAI has introduced a new thinking-time toggle for GPT-5 in ChatGPT. It is said to give more control over how quickly or deeply the model responds. The feature is now live but only available for Plus, Business, and Pro users. It appears in the message composer whenever you select GPT-5 with Thinking enabled.
By default, queries use Standard thinking time, which balances speed and intelligence. However, now with this feature, users will be able to switch to Extended for more thorough reasoning.
If you are a Pro user, you will get two extra settings, Light for the fastest responses and Heavy for the most comprehensive answers. Once you pick a setting, ChatGPT remembers your preference for future chats until you change it again.
According to the company, this update addresses feedback that GPT-5’s deeper reasoning mode sometimes felt slow for quick tasks. Light and Standard are intended for snappier replies without sacrificing accuracy, while Extended and Heavy give GPT-5 more time to process complex or high-stakes queries.
The update is not available on mobile yet. But OpenAI has confirmed that the toggle will arrive on its iOS and Android apps in the coming weeks.
Though the other day, OpenAI upgraded search to give more accurate results, reduce hallucinations, and better understand shopping-related queries. The company also refined factuality and reformatted answers to be easier to scan, improving the overall user experience.
One thing I have noticed after the launch of GPT-5 is that the AI giant is continuously launching new features or fixing some problems almost every day. And now with this new toggle, we can see that OpenAI is trying to balance between speed, accuracy, and usability.
