xiufengsun/TokenTracker

What it solves

TokenTracker provides a local-first dashboard to track token usage and costs across dozens of different AI coding tools. It eliminates the need to manually check multiple provider dashboards or manage API keys to understand how much AI resources are being consumed during development.

How it works

The tool uses a combination of auto-installed hooks, plugins, and passive readers to collect token data from 31 supported AI tools. It reads from local SQLite databases, JSONL logs, and configuration files produced by these tools. This data is then processed through a cost engine that uses LiteLLM pricing (refreshed daily) to calculate USD costs. All data remains local by default, though users can opt-in to a global leaderboard or cloud sync for cross-device views.

Who it’s for

Developers who use multiple AI coding assistants and want a unified, privacy-preserving way to monitor their token burn, costs, and usage trends without leaving their local environment.

Highlights

• Wide Compatibility: Supports 31 AI tools including Cursor, Claude Code, GitHub Copilot, and Gemini CLI.
• Privacy-First: 100% local data storage; no prompts, responses, or file contents are ever collected.
• Zero Config: Hooks and plugins are automatically installed on the first run.
• Gamified Experience: Includes a desktop pet that reacts to coding activity, 15 achievement tracks, and a global leaderboard.
• Native Integration: Offers macOS menu bar and Windows system tray apps with desktop widgets for real-time usage monitoring.