Demo data. No AI provider or gateway is connected yet — every figure on this screen is seeded sample data, not production telemetry.
Total cost
$6,140
last 60 days
Cost per successful task
$139.556
$130.648 per task
Success rate
93.6%
47 tasks
Retries
18
avg latency 7.4s
Potential savings
$1,658
Spend trend
Model usage
Spend by model for this agent
Recommendations
1 open
Enable prompt caching on ReplyDrafter preamble
A 6.1k-token static preamble is re-billed on every one of ~41k monthly calls.
Est. $1,240/month · quality 97.2% → 97.2%
- Open recommendations
Recent failures
3 failed tasks · $394 spent