Track AI usage, costs and plan limits — right from your desktop.
Usage, costs & plan limits.
One quiet place to keep track.
A quiet glance at your usage. Right at the notch.
Illustrative macOS demoShare the activity you choose. Find your people, compare usage and follow your progress.
| Rank | Builder | Tokens | Activity |
|---|---|---|---|
| 01 | N nova | 18.6M | |
| 02 | T tess | 15.2M | |
| 03 | K kai | 12.8M |
Compare capabilities, provider listings and pricing. See what is included in your plans.
A glance at the notch. A widget on your desktop. The full picture when you need it.
From your first request to your local inference stack, everything has a place.
Read the documentationUsage by tool and model, history, activity heatmaps and provider reset windows.
CPU, memory and I/O alongside MLX and Ollama telemetry, measured throughput and supervised inference engines.
An optional local gateway captures traces, timing and provider-reported usage. Request bodies stay on your machine.
Shell history, the th CLI, a local HTTP API and MCP tools bring the same view into your terminal and coding agent.
Usage tracking stays on your device. Publish a community profile only when you choose. Prompt sharing is off by default.
How your data is handled
Choose your desktop. Install the app.
Keep your usage in view.
No account needed for local tracking. Your data stays on your machine unless you choose to share.
The macOS app includes the notch and desktop widgets. Windows and Linux use the desktop dashboard.
Full installation guide
