Skip to main content

​
Overview

Reports are divided into several clusters, each focused on a specific aspect of your system’s performance.
You can use the dropdown at the top of the page to select a report cluster and filter your view by project and time range. Each report updates in real time, giving you immediate insight into how models and deployments are performing in production.

​
Customizing Reports

You can build your own custom report by selecting and combining charts from different clusters.
Use the time range and project filters to focus on the data that matters most to you. For example:
  • Compare total cost and latency trends over the past 30 days
  • Analyze error rate distribution for specific deployments
  • Examine which providers contribute most to overall usage and cost
Each visualization supports hover states, tooltips, and detailed metric breakdowns for deeper analysis.

​
Key Metrics and Charts

Custom Reports provide visibility into key performance and cost metrics, including:
  • Cost breakdowns across providers, models, and deployments
  • Latency distribution to identify slow responses
  • Response time by model for performance benchmarking
  • Error rates and retry frequency
  • Request and token volume across time and projects
These insights help teams identify bottlenecks, optimize spending, and track how improvements affect performance over time.