Cost Tracking
Chau7 shows your AI spending in dollars and cents. Per call, per session. No more surprises.
What is Cost Tracking in Chau7?
Cost Tracking is a feature in the Chau7 terminal that calculates the dollar cost of every LLM API call. Chau7 reads the token counts from each API response and multiplies them by the per-token price for the model used.
Chau7 displays costs at multiple levels: per individual API call, per run (a single agent invocation), and per session (all runs in a tab).
How to see dollar cost of AI agent API calls
Chau7 intercepts every LLM API call through its built-in TLS/WSS proxy. When the response comes back, Chau7 extracts the model name and token counts, then looks up the per-token price in its pricing table.
The dollar cost for each API call appears in Chau7's tab overlay. Cumulative costs for the entire session appear in the session view.
This hierarchical view in Chau7 lets developers find expensive operations at any granularity, from a single API call to a full session.
How does Chau7 know the pricing for each model?
Chau7 ships with a built-in pricing table that covers major LLM providers and models, including Claude 4.5, Claude 4.6, and other latest models. The pricing table includes separate rates for input tokens and output tokens.
Developers can override pricing for any model by placing a pricing.json file at ~/.chau7/pricing.json. Custom prices take precedence over the built-in table. The built-in pricing table is updated with each release to reflect current provider pricing from OpenAI, Anthropic, and Google.
How does Chau7's cost tracking compare to other terminals?
Most terminals do not track AI costs at all. iTerm2, Warp, Alacritty, and Kitty have no built-in cost tracking for LLM API calls.
Provider dashboards from OpenAI and Anthropic show only aggregate totals across all usage. They do not break costs down per session, per coding task, or per terminal tab.
Chau7 is the only terminal that calculates per-call dollar costs using a built-in pricing table with custom override support, then rolls those costs up per session. Chau7's cost tracking lets developers answer "how much did that refactor cost?" for any individual task.
Why cost tracking matters
Developers report AI coding costs exceeding five thousand dollars per month with no breakdown of where the money goes. Activity Monitor does not help. Provider dashboards show only aggregate numbers.
Chau7's cost tracking calculates cost per API call using the built-in pricing table with custom override support, then rolls costs up per session. Developers can see exactly how much each AI coding session costs as it happens.
Can I review costs through the MCP server?
Yes. Chau7 exposes run telemetry through its MCP server, which includes cost data per run. Developers can query this data to review costs programmatically.
Chau7's real-time cost display in the tab overlay makes it easy to notice when a session is becoming expensive.
Questions this answers
- What is Cost Tracking in Chau7 terminal?
- How does Chau7's cost tracking compare to other terminals?
- How to see dollar cost of AI agent API calls?
- How does Chau7 know the pricing for each model?
- Can I review costs through the MCP server?
Frequently asked questions
What is Cost Tracking in Chau7 terminal?
Cost Tracking is a feature in the Chau7 terminal that calculates the dollar cost of every LLM API call by multiplying token counts by the built-in pricing table. Chau7 displays costs per call and per session so developers can see exactly where their AI budget goes.
How does Chau7's cost tracking compare to other terminals?
Most terminals do not track AI costs at all. iTerm2, Warp, Alacritty, and Kitty have no built-in cost tracking. Provider dashboards show only aggregate totals. Chau7 is the only terminal that calculates per-call dollar costs using a built-in pricing table, then rolls them up per session.
How to see dollar cost of AI agent API calls?
Chau7 intercepts every LLM API call through its TLS/WSS proxy, reads the token counts from the response, and multiplies them by the per-token price for that model. The dollar cost appears in Chau7's tab overlay and session view.
How does Chau7 know the pricing for each model?
Chau7 ships with a built-in pricing table for major providers and models including Claude 4.5, Claude 4.6, and other latest models, with different rates for input and output tokens. Developers can override pricing for any model by placing a pricing.json file at ~/.chau7/pricing.json. Custom prices take precedence over the built-in table.
Can I review costs through the MCP server?
Yes. Chau7 exposes run telemetry through its MCP server, which includes cost data per run. Developers can query this data to review costs programmatically.
Does cost tracking work with self-hosted or custom models?
Yes. Developers can override pricing for any model by placing a pricing.json file at ~/.chau7/pricing.json. Custom prices take precedence over the built-in table, so self-hosted or custom models can have accurate cost tracking. The built-in table covers major providers (OpenAI, Anthropic, Google) including Claude 4.5 and Claude 4.6.