Do you sometimes feel that the AI answers you at lightning speed, like a student in a rush to get their homework over with? Or conversely, do you get bored waiting for it to solve a very simple problem that shouldn’t require all that effort? If you use ChatGPT regularly, you might have missed a wonderful and very important feature that allows you to precisely control the “depth of thinking” of the model before it gives you the final answer. This feature completely changes how we interact with AI and saves a lot of time and effort, and the difference between the various settings might truly surprise you!
What is the “Reasoning Effort” feature and how does it work?
In advanced models like o1 and o3-mini, OpenAI introduced a new concept known as “Reasoning Chains.” Instead of generating the next words directly in a statistical manner, these models take extra seconds for internal reflection, reviewing steps, and correcting errors before showing the answer to the user. However, sometimes you may not need all this “philosophical depth” to write a simple email or translate a short sentence.
This is where the “Reasoning Effort” setting comes in. This feature allows you to manually choose between multiple levels: Low, Medium, or High. By lowering the effort, you can get extremely fast answers similar to the quick GPT-4o model, while raising the effort to the maximum gives you the full power of the model to solve the most complex programming and mathematical problems that require superior focus.
How do you access this secret setting in ChatGPT?
Accessing this feature is not complicated, but it requires you to know where to look. When you open a new chat in ChatGPT (whether via the iPhone app or the web) and select a model that supports advanced reasoning like o1, you will find a new option appearing next to the model name called “Reasoning.” When you click on it, a dropdown menu will appear, allowing you to drag the slider or choose the desired effort level.
Source:
Leave a Reply