> See the full index of documentation pages at https://docs.zettabyte.com/llms.txt.

# Token usage

> Track inference token consumption and cost.

Source: https://docs.zettabyte.com/docs/deploy-and-serve/token-usage

Last updated: 2026-07-17

Token Statistics, in the zSTUDIO workspace, reports inference token consumption over a date range you select. Calls to a deployed endpoint over the inference API consume tokens, which are metered here. Summary cards show Total Token Usage, Input Tokens, Output Tokens, and Total Expense.

Below the cards, a table breaks the figures down two ways:

- By API Key: usage per key, so you can attribute consumption to a specific key.
- By Model: usage per model.

Each row shows Input Tokens, Output Tokens, Total Tokens, and Cost. For how token cost fits into billing, see How billing works.