What happened
Over the last two or three updates, the quota limit system has changed significantly; it shifted from a simple one-hour interval between sessions to a token-based usage system. To make matters worse, the models now appear to be slow—even in "Low (slower)" mode, it takes between 30 and 70 seconds for a skill to start executing, which wastes your usage time and makes you slightly less productive.
Steps to reproduce
There is no steps to reproduce
Where does this happen?
Desktop app
Operating system
Windows
Version
Latest 0.94
Model
GLM 5.3 Flash
Logs or screenshots
What happened
Over the last two or three updates, the quota limit system has changed significantly; it shifted from a simple one-hour interval between sessions to a token-based usage system. To make matters worse, the models now appear to be slow—even in "Low (slower)" mode, it takes between 30 and 70 seconds for a skill to start executing, which wastes your usage time and makes you slightly less productive.
Steps to reproduce
There is no steps to reproduce
Where does this happen?
Desktop app
Operating system
Windows
Version
Latest 0.94
Model
GLM 5.3 Flash
Logs or screenshots