Skip to main content
The LLM Activity Dashboard gives you visibility into how AI coding agents are using LLM APIs across your organization. Navigate to Monitor > Dashboards and select the LLM Activity tab. Use the 24h / 7d / 30d toggle in the top right to select the time range (30 days is shown by default).
Turen LLM Activity dashboard with 45,312 total requests, 2.3s average latency, 0.4% error rate, a provider donut chart, and a requests-by-model bar chart.

The LLM Activity dashboard showing total requests, latency, error rate, data volume, active devices, and requests broken down by provider and model.

​
Key Metrics

The dashboard summarizes activity across a set of headline metrics:

​
Provider Breakdown

A donut chart showing how usage is distributed across LLM providers:
  • Anthropic: Claude models
  • OpenAI: GPT models
  • Google: Gemini models
  • Mistral: Mistral models
  • Cohere: Cohere models

​
Model Breakdown

A horizontal bar chart showing the models your team is using, ranked by request volume. Model names are displayed in a friendly format (e.g., “Claude Opus 4.6”, “GPT-5.2 Pro”). This helps you:
  • Track adoption of newer models
  • Understand usage distribution
  • Identify usage patterns across your team

​
Session Activity

The bottom section shows detailed session metrics:
  • Total Sessions: Number of Claude Code sessions in the period
  • Total Messages: Messages exchanged across all sessions, with daily average
  • Avg Duration: Average session length
  • Unique Clients: Number of distinct machines with sessions
  • Sessions This Week: Daily session count bar chart
  • Activity by Hour: 24-cell heatmap showing when your team is most active
  • Top Clients: Ranked list of machines with the most session activity, showing hostname, session count, total bytes, and last active time