Cost & Latency Dashboards
FinOps includes two prebuilt dashboards — Cost and Latency — that surface the most-used spend and performance views without any setup.
ℹ
Open them from FinOps → Cost Dashboard and FinOps → Latency Dashboard. Want a custom view instead? Build one with Custom Dashboards.
Cost Dashboard
A model-level view of spend:
- Cost by model — which models drive your bill.
- Total spend over the period.
- Volume — traces/requests over time.
- Cost by user — spend attributed to end users.
- Score trends — quality alongside spend, so cuts don't quietly hurt quality.
Latency Dashboard
A performance view:
- Latency distribution — p95 / p99 response times.
- Generation latency by model — which models are slow.
- Score analytics — quality across agents alongside latency.
Filters
Both dashboards support date range, environment, and custom trace filters.
When to use which
- Use the Cost Dashboard for budget reviews and finding spend hotspots.
- Use the Latency Dashboard for performance/SLO reviews.
- For the cost-versus-quality trade-off specifically, see Cost & Performance Metrics.
Next steps
- Cost & Performance Metrics — cost vs. quality, in depth.
- AI Optimizer — act on what the dashboards reveal.