Skip to main content
Use the Usage page in the Brightnode console to monitor hosted inference activity across your organization.

​
What you can track

Usage analytics is focused on inference traffic and includes:
  • Request volume over time.
  • Input and output token usage.
  • Cost trends.
  • Latency percentiles.
  • Top models by traffic and spend.
The console currently supports 24h, 7d, and 30d views so you can inspect recent spikes or look at broader trends.

​
Why it matters

Usage analytics helps you answer practical questions such as:
  • Which models are driving most of your spend?
  • Are latency patterns changing after a model switch?
  • Is traffic growing in a predictable way?
  • Which integrations should move to a different model or budget tier?
  1. Create separate API keys for each environment or service.
  2. Route traffic through Router with explicit model IDs.
  3. Review usage regularly to spot cost, performance, or adoption changes.
If you need dedicated managed inference instead of the shared router path, see Beams.