Skip to content

Usage Tab

The Usage tab provides comprehensive statistics and analytics for your AI Text Editor usage, helping you track LLM API calls, token consumption, and performance metrics.

The Usage tab is located in the AI sidebar and offers detailed insights into:

  • API call statistics
  • Token usage patterns
  • Performance metrics
  • Session-based analytics
  • Provider and model usage

Features

Usage Statistics Dashboard

The main dashboard displays key metrics:

  • Total Calls: Number of API calls made
  • Total Tokens: Combined input and output tokens
  • Input Tokens: Tokens sent to the LLM
  • Output Tokens: Tokens received from the LLM
  • Total Time: Cumulative processing time
  • Average Time: Mean response time per call

Filtering Options

Filter your usage data by:

  • Date Range: All time, today, last 7 days, last 30 days, or custom range
  • Session: Current session, all sessions, or specific session IDs
  • Provider: Filter by AI service provider (Anthropic, OpenAI, Groq, etc.)
  • Model: Filter by specific model within a provider

Usage Subtabs

Overview

Displays the main statistics dashboard with filtered data.

By Prompt

Shows detailed usage statistics grouped by prompt:

  • Call count per prompt
  • Token consumption breakdown
  • Execution time metrics
  • Last usage timestamp
  • Session participation (when viewing all sessions)

Recent Calls

Lists individual API calls with:

  • Prompt identifier
  • Timestamp and relative time
  • Token usage details
  • Provider and model information
  • Session identification
  • Search and limit controls

Controls and Actions

Refresh

  • Refresh Button: Manual data refresh
  • Auto-refresh: Automatic updates every 30 seconds when tab is active

Export

  • Export Data: Download complete usage history as JSON
  • Includes all calls, tokens, timing, and metadata
  • Filename format: llm-usage-export-YYYY-MM-DD.json

Clear History

  • Clear Usage: Remove all stored usage data
  • Requires confirmation dialog
  • Action cannot be undone

Data Storage

Usage data is stored locally using IndexedDB for:

  • Persistent storage across browser sessions
  • Efficient querying and filtering
  • Large dataset support
  • No server dependencies

Search and Navigation

Search Functionality

  • Filter calls by prompt name or session ID
  • Real-time search results
  • Case-insensitive matching

Pagination

  • Configurable result limits (50, 100, 500, or all)
  • Efficient loading for large datasets
  • Responsive performance

Session Integration

The Usage tab integrates with the Session Manager to:

  • Track calls per session
  • Identify current vs. historical sessions
  • Provide session-based analytics
  • Enable cross-session comparisons

Performance Metrics

Duration Tracking

  • Response time measurement in milliseconds
  • Automatic formatting (ms, seconds, minutes, hours)
  • Average calculation across filtered datasets
  • Performance trend analysis

Token Economics

  • Input/output token breakdown
  • Total consumption tracking
  • Provider-specific metrics
  • Cost estimation support (when available)

Best Practices

Monitoring Usage

  1. Regular review of token consumption
  2. Identify high-usage prompts
  3. Monitor performance trends
  4. Track session efficiency

Data Management

  1. Periodic export for backup
  2. Clear old data when needed
  3. Monitor storage usage
  4. Review session patterns

Optimization

  1. Analyze prompt efficiency
  2. Compare provider performance
  3. Optimize high-usage workflows
  4. Balance token usage across prompts