Tokens and USD per design, model (the OpenRouter id actually used, not the gateway alias), and agent role. Cost prefers LiteLLM response_cost, then the llm.prices table.
Totals
Calls
–
Designs
–
Tokens
–
Cost
–
Prompt
–
Completion
–
Reasoning
–
Avg latency
–
Averages per design
Tokens / design
–
Cost / design
–
Calls / design
–
Per design
Each ChatCraft design
| Key | Calls | Tokens | Prompt | Completion | Reasoning | Cost | Avg tokens | Avg cost | Avg latency |
|---|---|---|---|---|---|---|---|---|---|
| No LLM usage in this window. | |||||||||
Per model
Actual model id, not the furn-* alias
| Key | Calls | Tokens | Prompt | Completion | Reasoning | Cost | Avg tokens | Avg cost | Avg latency |
|---|---|---|---|---|---|---|---|---|---|
| No LLM usage in this window. | |||||||||
Per role
Agent that made the call
| Key | Calls | Tokens | Prompt | Completion | Reasoning | Cost | Avg tokens | Avg cost | Avg latency |
|---|---|---|---|---|---|---|---|---|---|
| No LLM usage in this window. | |||||||||