OpenAI has quietly rolled out new hidden features in the ChatGPT app, allowing users to control the intelligence level of their prompts. Alongside these updates, users may also notice changes in response quality, speed, reasoning ability, and overall behaviour, as the default model has been updated recently.
A key new addition is a gesture-based control that lets users adjust how much effort ChatGPT should apply before generating a response. By pressing and holding the send arrow, users can open an intelligence-level selector.
This feature allows users to choose between different modes such as Instant, Thinking, and Extended, depending on how detailed or fast they want the response to be. However, availability may vary based on the user’s subscription plan.
OpenAI has also introduced a table of contents feature to make long conversations easier to navigate. When a chat includes five or more responses, ChatGPT automatically organizes the discussion into sections based on topics.
A menu icon appears on the right side of longer conversations, and hovering over it reveals the table of contents, allowing users to quickly jump between different parts of the chat. At present, this feature is available only on the web version of ChatGPT.
In addition, OpenAI has made improvements to the GPT-5.5 Instant model, aimed at enhancing response style and quality. The company says these updates are intended to deliver a more refined and consistent user experience.
Overall, these updates indicate a shift toward giving users more control over how ChatGPT thinks, responds, and organizes information during conversations.
Recent Random Post:
