Cost Tracking
Your AI spending in dollars and cents. Per call, per session, per month. No more surprises.
Questions this answers
- How much does Claude Code cost per session?
- LLM API cost monitor dashboard for terminal
- Track AI coding agent spending in real time
- How to see dollar cost of AI agent API calls?
- Why is my AI coding bill so high?
How it works
Cost Tracking takes the token counts from each API call and multiplies them by the per-token pricing for the model used. It maintains an up-to-date pricing table for major providers and models, distinguishing between input and output token rates. The result is a dollar cost for every single API call your AI agent makes.
Costs roll up at multiple levels. You can see the cost of an individual API call, the total cost of a run (a single agent invocation), the cumulative cost of a session (all runs in a tab), or the global cost across all tabs and sessions. This hierarchical view lets you find expensive operations at any granularity.
For teams and heavy users, cost trends over time reveal whether spending is stable, growing, or spiking. Combined with the Analytics Dashboard, you get a clear picture of where your AI budget goes.
Why it matters
Developers report AI coding costs exceeding five thousand dollars per month with no breakdown of where the money goes. Activity Monitor won't help. Your provider dashboard shows aggregate numbers. Chau7 calculates cost per API call using current provider pricing tables, then rolls it up per session and per period. You'll finally be able to answer 'how much did that refactor cost?'
Frequently asked questions
How does Chau7 know the pricing for each model?
Chau7 ships with a built-in pricing table for major providers and models, including different rates for input and output tokens. The table is updated with each release to reflect current pricing.
Can I set a cost alert or budget limit?
Budget alerts and spending limits are planned for a future release. Currently, you can monitor costs in real time and review cumulative spending through the MCP server and Analytics Dashboard.
Does cost tracking work with self-hosted or custom models?
If the API response includes model name and token counts, Chau7 can calculate costs as long as the model's pricing is configured. You can add custom model pricing through the settings.