Usage analytics: track AI coding costs and token usage
CodexUse 1.0.3 adds a full analytics dashboard. See exactly how many tokens you're using, what it's costing, and which profiles are most active.
Why track usage?
AI coding assistants don't show you spending in real-time. You find out at the end of the month when the bill arrives. With multiple accounts, it's even harder to know where tokens are going.
The analytics dashboard solves this by tracking every conversation automatically:
- Token counts - Input, output, and cache tokens per session
- Cost estimates - Based on current model pricing
- Profile breakdown - See which accounts use the most
- Model usage - Know which models consume your quota
What's in the dashboard
Usage summary cards
At the top, four cards show your key metrics for the selected time range:
- Total tokens used
- Estimated cost
- Number of sessions
- Active profiles
Usage trend chart
A line chart shows token usage and cost over time. Toggle between daily and hourly views. Useful for spotting usage spikes or tracking a busy project week.
Model breakdown
A pie chart shows which models you're using most. Each slice shows token count and percentage. Helps you understand if you're using expensive models when cheaper ones would work.
Profile comparison
A bar chart compares all your profiles by cost, tokens, or session count. The leaderboard ranks profiles by activity. See which accounts are doing the heavy lifting.
Activity heatmap
A calendar heatmap shows your coding activity patterns. Darker cells mean more usage. Useful for understanding your own habits.
Session stats
A table lists recent sessions with thread name, token count, cost, and timestamp. Click to see details about any conversation.
Date range filtering
Use the date picker to filter analytics:
- Today - Current day only
- Last 7 days - Past week
- Last 30 days - Past month
- Custom - Pick any start and end date
All charts and summaries update when you change the range.
Pro feature
Analytics is a Pro feature. Free users see the dashboard but need to upgrade for full data. Pro costs $19 one-time and includes unlimited profiles, auto-roll, and rate-limit monitoring.
Data storage
All analytics data is stored locally in your ~/.codex directory. Nothing is sent to external servers. CodexUse reads token counts from CLI responses and calculates costs based on published model pricing.
Related
Does CodexUse track my AI usage automatically?
Yes. CodexUse automatically tracks tokens, costs, and sessions for every conversation across all your profiles. Data is stored locally.
Can I compare usage across different AI accounts?
Yes. The profile comparison chart shows cost and token usage side-by-side for all profiles. A leaderboard ranks profiles by activity.
What time ranges can I view in analytics?
You can filter by today, last 7 days, last 30 days, or a custom date range. Switch between daily and hourly views for trends.