Skip to main content

Usage & Cost

The Usage & Cost page provides comprehensive analytics into API throughput, token consumption, failure rates, and infrastructure spending across your organization.

Navigate to OBSERVABILITY → Usage in the Aegis Console to inspect global metrics or slice metrics by teams, agents, virtual keys, or runtimes.

Key Metrics Summary (KPI Cards)​

At the top of the Usage page, five key operational metrics provide an immediate snapshot of platform activity:

MetricDescription
Total SpendTotal estimated consumption cost in USD ($) for the selected timeframe.
Total RequestsTotal volume of API requests processed by the gateway.
Total TokensTotal token volume, broken down by Prompt Tokens and Completion Tokens.
Fail Rate (%)Percentage of failed API calls. Color-coded for system health (<1% Green, <5% Yellow, ≥5% Red).
Avg LatencyAverage end-to-end processing latency across requests in milliseconds (ms).

Usage Over Time Chart​

The main chart visualizes platform activity trends over time.

Chart Metric Toggle​

  • Requests & Errors: Displays line series for total API request volume vs failed request counts (Errors).
  • Tokens: Displays stacked bar charts showing Prompt Tokens (input) and Completion Tokens (output) per time bucket.

Automatic Time Bucketing​

The chart automatically adjusts time bucket intervals based on the selected date range:

  • ≤ 2 hours: 5-minute buckets (5m)
  • ≤ 12 hours: 15-minute buckets (15m)
  • ≤ 48 hours: 1-hour buckets (1h)
  • ≤ 14 days: 6-hour buckets (6h)
  • > 14 days: 1-day buckets (1d)

Usage Breakdown Tables​

Click "Show Usage Breakdown" to view tabular breakdowns sliced by specific platform dimensions:

Dimensions​

  • By Teams: Aggregates usage and spend by team workspace.
  • By Agents: Aggregates usage and spend per AI Agent.
  • By Keys: Aggregates usage per API Virtual Key.
  • By Runtimes: Aggregates usage across connected agent execution runtimes.

Breakdown Table Schema​

ColumnDescription
NameName/Alias of the team, agent, key, or runtime.
RequestsTotal request count handled by the entity.
Prompt TokensTotal prompt (input) tokens processed.
Completion TokensTotal completion (output) tokens generated.
Total TokensSum of prompt and completion tokens.
Spend (USD)Total cost accrued by the entity in USD ($).
Fail RatePercentage of requests resulting in errors for that specific entity.

Date Range Selection​

Use the Date Range Picker in the top right to filter all metrics:

  • Quick presets: 1 hour, 24 hours, 7 days, 30 days.
  • Custom absolute start and end date/time range.