OpenAI has rolled out a new feature for ChatGPT, giving users more control over how long the chatbot “thinks” before delivering a response. The update addresses long-standing complaints about GPT-5 and its reasoning-focused counterpart, GPT-5 Thinking, which often took too long to answer simple queries.
ChatGPT Gets a Thinking Duration Toggle
In a post on X (formerly Twitter), OpenAI announced a new toggle that allows users to adjust ChatGPT’s thinking duration. The options include Light, Standard, Extended, and Heavy — with Light generating responses fastest and Heavy allowing the model to take the longest time for more complex reasoning.
Currently, the feature is limited by subscription tiers. Free and ChatGPT Go users do not have access to GPT-5 Thinking. Plus and Business subscribers can use Standard and Extended, while Pro users gain access to all four modes.
Limitations and Rollout
At launch, the toggle is only available on the ChatGPT website, with desktop and mobile app support expected later. OpenAI noted the update came directly from user feedback about the lack of control over response timing.
“This new feature is part of our effort to give users more control and improve the overall ChatGPT experience.”
OpenAI (via X)
More Updates Coming
OpenAI has also been updating personalization options in ChatGPT, including a new settings page that integrates personality, custom instructions, and memory features in one place. According to OpenAI’s official blog, these features are expected to roll out globally soon.
Related Reads
- OpenAI Blog: Latest Product Announcements
- Follow OpenAI on X (Twitter)
- TechCrunch Coverage of ChatGPT
- https://t.co/8NqoMNT8ee