Manually Scale AI Reasoning Effort for Complex Logical Tasks
Adjust the reasoning effort level based on task complexity.
Use the updated model picker (Instant to Pro Extended) to manually select the 'reasoning effort' required for your specific task to balance speed and accuracy.
The Scenario
You are working on a complex strategic document or a difficult debugging problem that requires deep logical reasoning rather than just a quick summary.
Before & after
Previously, you might have had to prompt the AI multiple times or use different tools to get deep reasoning, often spending 10–15 minutes refining the output.
By selecting 'Medium', 'High', or 'Extra High' in the new model picker, you can force the AI to spend more time 'thinking' to solve the logic, taking about 30–60 seconds to generate a high-quality answer.
The Prompt
[SET_PICKER_TO_HIGH] Analyze this technical architecture for potential security bottlenecks and suggest a three-phase mitigation plan: [PASTE_ARCHITECTURE_DETAILS]
The updated model picker replaces 'Thinking' labels with a clearer hierarchy: Instant, Medium, High, Extra High, Pro Standard, and Pro Extended. Use 'Instant' for quick drafts and 'High' or 'Extra High' for complex strategy or coding.
Source
ChatGPT Business - Release Notes | OpenAI Help Center"makes it easier to choose the balance of speed and reasoning effort that works best for a task"