TokenBar tracks AI token consumption and spend in real time from the macOS menu bar, providing an ambient cost indicator for developers and heavy AI tool users who want visibility into their usage without logging into provider dashboards.
The menu bar display shows the current session burn rate, daily token total, remaining credits, and upcoming reset date for each connected provider. Supported providers include OpenAI (including Codex), Anthropic Claude, Google Gemini, and Cursor. Users configure cost thresholds and receive alerts when approaching them, avoiding unexpected charges at the end of a billing period. The app is designed to operate like a battery indicator—glanceable and persistent—rather than a tool that requires active interaction.
All data processing happens locally on the user’s Mac using credentials supplied by the user to query provider usage APIs. No usage data is transmitted to TokenBar’s servers. The Basic tier ($5 one-time) covers live usage visibility for standard provider accounts, while the Pro tier ($10 one-time) adds expanded support for more Cursor workflows and additional tracking granularity.
System requirements: macOS 14.0 Sonoma or later. Compatible with Apple Silicon and Intel Macs. Available from the Mac App Store (App Store ID 6760401627) and direct download at tokenbar.site.
Pricing: ($5 Basic or $10 Pro, one-time purchase) No subscription.
Limitations: Tracking is limited to the specific AI tools and providers supported by TokenBar; usage in browser-based AI chat interfaces not connected via API is not captured. The app depends on provider APIs being available and consistent; changes to a provider’s usage reporting format may temporarily break tracking until an update is released.
Alternatives: AI Cost Bar (free alternative for LLM cost monitoring, browser-based); provider dashboards (free, accurate but require manual navigation and don’t provide real-time ambient visibility); open-source ai-token-monitor on GitHub (Claude Code–focused, free, requires technical setup).
Suitable for developers running frequent AI-assisted coding sessions who want a persistent cost gauge across multiple providers without switching between browser tabs to check usage.
