Model Usage

AIモデル

AIモデルのコストとトークン使用量を時系列で追跡

できること

  • Per-model cost breakdown — See exactly what each model costs in your sessions
  • Current model usage — Most recent session's model and cost details
  • Full model breakdown — Complete historical cost analysis across all models
  • Multiple providers — Track usage for Codex and Claude Code
  • Flexible input — Read from CodexBar CLI directly, from file, or stdin
  • Output formats — Text for readability or JSON for scripting

    • 試してみる質問

  • "Show my current Codex session's model usage and cost"
  • "Break down costs across all Claude models I've used"
  • "What's my total spend for today?"
  • "Export usage data as JSON"
  • "Show token counts for each model"

    • プロのコツ

  • JSON output is perfect for expense tracking and budget monitoring
  • Regular monitoring prevents surprise bills from expensive models
  • Compare model costs to make informed choices about which to use
  • File input enables historical analysis across multiple sessions
  • Token counts help identify opportunities to optimize prompt efficiency