My Usage
The My Usage page shows your personal AI coding usage analytics.
Token & cost cards
Section titled “Token & cost cards”At the top, summary cards show:
- Total tokens — input + output tokens across all sessions
- Estimated cost — calculated from token counts and model pricing
- Sessions — number of coding sessions in the selected period
- Average session length — in turns
Date range
Section titled “Date range”Use the date picker to filter by time period. Defaults to the last 7 days. Options include:
- Last 24 hours
- Last 7 days
- Last 30 days
- Custom range
Model mix
Section titled “Model mix”A breakdown of token usage by model:
| Model | Typical use | Cost tier |
|---|---|---|
| Opus | Complex reasoning, architecture | Highest |
| Sonnet | General coding, most tasks | Medium |
| Haiku | Simple tasks, formatting | Lowest |
The model mix chart helps identify model overkill — using Opus for tasks Sonnet could handle.
Daily trends
Section titled “Daily trends”A line chart showing daily token consumption and cost. Useful for spotting:
- Spikes from long sessions or complex tasks
- Trends toward more or less efficient usage
- Impact of changing your workflow (e.g., using plan mode more)
Efficiency metrics
Section titled “Efficiency metrics”- Tokens per turn — lower is better (concise, focused prompts)
- Cost per session — average session cost
- Output/input ratio — how much output you get per input token