Max effort isn't the best effort (the 2-step test)
Max effort can cost about 13 times more than low and still not give the best result. Run your real task at medium, then once at high, and keep high only if it's clearly better.
The prompt
The 2-step effort test (for one real task you do often): 1. Run the task at medium effort, the default. [paste your real task here] 2. Run the exact same task once at high effort. 3. Put the two results side by side. - Same result? Stay at medium. - Clearly better? Keep high for this kind of task. Max is not a quality button. In one public test, max cost about 13 times more than low, and it didn't build the best app; the level one step below max did.
How to use it
- Pick one real task you run often, and run it at medium effort, the default.
- Run the exact same task once at high effort, and put the two results side by side.
- Same result? Stay at medium. Clearly better? Keep high. Don't reach for max by default.