Skip to main content
Global AI records the details of every API call and supports filtering by time, model, token, and more. Click Usage logs in the left sidebar, or go directly to https://globalai.vip/usage-logs/common to see your call records.

View call logs

Each row in the log list is one call and includes:
  • Call time
  • Model used
  • Tokens consumed (prompt tokens, completion tokens, and cache tokens)
  • Quota consumed
  • Call status and latency
  • Token name used
Click the detail icon on the far right to see more information about the call.

Search and filter

1

Expand the filter panel

Click Expand at the top of the Usage logs page to reveal the filter area.
2

Set filters

Common options:
  • Date range: choose a start and end date
  • Model: enter a keyword such as gpt-5.6-sol
  • Token name: select or type a token name
  • Type: consumption only or top-ups only
3

Search

Click Search. The list refreshes with the filtered results.
Filtering by token name is the fastest way to trace a specific application’s consumption. Remember to use meaningful names when you create a token.

Dashboard

1

Open the dashboard

Click Dashboard in the left sidebar, or go directly to https://globalai.vip/dashboard/models.
2

Review the charts

The page uses line or bar charts to show daily API call volume and quota consumption trends. Hover over a chart to see values for a specific date.
3

Switch dimensions

Switch views by model, token, or time granularity to quickly identify the fastest-growing model or the token spending the most quota.

Common analysis scenarios

In the dashboard, break down by token to find the top spender. Then filter the logs by that token to see which model or call pattern is responsible.
Filter the logs by token name and check whether call times and source IPs look unusual. If anything is off, immediately delete and recreate the token in Token management.
In the dashboard, view the last 30 days, note the daily consumption average and trend, then multiply by the number of active days you plan for. Actual cost also depends on the model multiplier; see Pricing.