TokenID captures every API call, detects overcharging, and shows you exactly where your money goes. Local-first. No signup. Free.
Python 3.10+ required. Install via pip:
pip install tokenidStart TokenID as a local proxy. It intercepts your LLM API calls and captures every token:
tokenid proxy --port 8787Set your LLM client's base URL to http://localhost:8787 and use your normal API key. TokenID forwards requests transparently.
Open the dashboard and see exactly how much you're spending — and overpaying:
tokenid dashboardLocal dashboard at http://localhost:8788. No data leaves your machine unless you opt into the free cloud tier.
Want your usage tracked across machines and a weekly spend report? Activate the free cloud tier:
Run the verify command. You'll get a one-click verification link:
tokenid verify you@company.comEvery Monday, TokenID emails you a breakdown: tokens consumed, models used, estimated cost, and detected overcharges.