Usage Metrics

The Token Analytics section provides a real-time overview of your resource consumption across four key metrics:

1. Prompt Tokens

• Measures tokens used in user inputs and system prompts
• Shows how much input text you’re sending to AI models
• Helps optimize prompt engineering for cost efficiency

2. Completion Tokens

• Measures tokens generated in agent responses
• Tracks output length and response generation costs
• Monitors AI-generated content volume

3. Total Tokens

• Combined count of prompt and completion tokens
• Provides overall consumption visibility
• Calculates total API usage costs

4. Request Counts

• Number of API calls made to AI models
• Tracks interaction frequency across agents
• Helps identify high-usage periods and patterns

Time Period Analysis

• Flexible Date Ranges: View usage data for specific periods (e.g., Jan 01, 2026 - Jan 14, 2026)
• Custom Filters: Filter by model type, agent, project, or specific time frames
• Historical Tracking: Compare usage across different periods to identify trends
• Real-Time Updates: Monitor consumption as it happens

Fig 10.12