Skip to content

Usage Reports

Usage reports provide time-series and hierarchical views of token consumption across your organization.

The time-series chart shows token consumption over time. Each data point represents one day (or one hour for the 24h range), stacked or grouped by model.

Selecting a date range:

PresetGranularity
Last 24 hoursHourly buckets
Last 7 daysDaily buckets
Last 30 daysDaily buckets
Last 90 daysDaily buckets

The chart displays separate series for each model used in the period, allowing you to see which models are driving consumption.

The hierarchy table shows cumulative token totals organized as:

Organization
└── Department
└── Project

Each level shows prompt tokens, completion tokens, and total tokens. Click a row to expand it and see the next level of detail.

This view is useful for:

  • Identifying high-consumption departments or projects
  • Chargeback reporting
  • Quota planning

You can filter the usage data by:

  • Date range — select a preset or custom range
  • Model — filter to a specific model or provider

Usage data is also available via API for programmatic reporting:

Time-series:

Terminal window
GET /api/usage/timeseries?start=2024-01-01&end=2024-01-31

Response:

[
{
"date": "2024-01-01",
"model_id": "llama3:8b",
"provider": "ollama",
"prompt_tokens": 12400,
"completion_tokens": 8200,
"total_tokens": 20600
}
]

Hierarchy:

Terminal window
GET /api/usage/hierarchy?start=2024-01-01&end=2024-01-31

Response:

[
{
"org_id": "...",
"org_name": "Acme Corp",
"total_tokens": 1250000,
"departments": [
{
"dept_name": "Engineering",
"total_tokens": 980000,
"projects": [...]
}
]
}
]