Plenty of frustrating sessions aren’t about intelligence at all. They’re about limits nobody mentioned.
Know these numbers for your tools
- Context window: how much text the model can consider at once. Past it, earlier details get dropped or summarized.
- Usage limits: requests or tokens per hour, day or month, and when they reset.
- Feature tiers: which models, tools and integrations your plan includes.
- Data policies: what your provider keeps and whether it’s used for training, especially on work accounts.
Five minutes with your plan’s documentation saves hours of wondering why the assistant “forgot” the file you shared an hour ago.