Think4Ever
Sign in

Usage & Analysis

The Usage & Analytics section provides a detailed breakdown of your platform utilization, token consumption, and transactional metrics. This screen helps you audit costs and understand exactly how your AI agents are consuming system resources.

Usage & AnalysisClick to enlarge

Sub-Tab Navigation Bar#

Located at the top of the workspace, these tabs allow you to toggle between different analytical views:

  • Overview (Current View): A high-level dashboard summarizing volume metrics and overall token balances.
  • By Model: Breaks down consumption data based on the specific AI model engines processing your requests.
  • By Model: Breaks down consumption data based on the specific AI model engines processing your requests.
  • Cost Breakdown: Translates token usage into actual financial expenditures based on your billing tier.
  • Projects: Segregates usage metrics by individual active workspaces or development branches.

High-Level Summary Cards#

This interactive data table itemizes execution metrics segmented by individual development environments inside the selected tracking window:

Daily Usage Time-Series Chart#

The top metrics section highlights cumulative totals for quick monitoring:

  • Total Requests: The aggregate number of execution calls or tasks initiated across your pipelines (e.g., 797 requests).
  • Total Tokens: The combined volume of processing units used by the system (e.g., 19.5M tokens). This includes input, output, and cached context.
  • Prompt Tokens: The standalone volume of input data sent to the models (e.g., 18.3M tokens), indicating the heavy baseline processing of instructions, documentation, and system state rules.

Token Breakdown Module#

This card explicitly itemizes how your total tokens are distributed during execution loops:

Token TypeMetric VolumeFunctional Purpose
Prompt tokens18.3MRepresents the raw input instructions, system architectures, and engineering requirements parsed by the AI models.
Completion tokens1.2MRepresents the actual output generated by your AI agents (e.g., code, generated text, or processed system steps).
Cache read tokens10.9MRepresents previously processed context that was successfully retrieved from memory. High values here mean the system is optimizing efficiency and saving costs by avoiding redundant processing.
Cache write tokens381.6KTracks new, reusable contextual data written into the platform's cache infrastructure to accelerate subsequent agent tasks.