How we calculate AI usage
We track AI activity in tokens behind the scenes, but we show usage to customers as simple usage points. This keeps plan limits easier to understand without changing how the system measures actual AI consumption.
Step 1
Every AI interaction generates token usage based on the prompt you send and the response you receive.
Step 2
We total those tokens in the system exactly as they are reported by the model provider.
Step 3
For customer-facing plan usage, we convert tokens into points using a simple formula: 1 point = 1,000 tokens.
The formula
2,500 tokens
Displays as 2.5 usage points
10,000 tokens
Displays as 10 usage points
500,000 tokens
Displays as 500 usage points
Why we show points instead of raw tokens
Raw token counts can be difficult to interpret at a glance. Showing points makes plan usage clearer, easier to compare across plans, and simpler for teams to understand while the backend still tracks exact token consumption for system accuracy.