Skip to main content

​
Per Response

Every envelope carries usage: prompt, completion, reasoning, and cached token counts, plus server_tool_use counters when web tools ran. Aggregate them in your own logs.

​
Per-Model Spend

The developer console at app.trynito.ai shows spend by model, daily totals, and request counts. It is a browser surface.

​
Balance reads

Use GET /api/v1/billing/balance with administrative credentials for the current balance. It does not return request history or total historical spend. A public billing-usage history endpoint is not yet available.