Running costs
Does It Cover Its Own Costs?
Every search, model call and logo is a paid request. This page sets what the agent spends on thinking against what the treasury pays back from fees: a fixed 0.0005 ETH per launch.
Paid by the operator for now. The agent is not running yet, so the costs below are an illustrative week. The API accounts behind the agent are paid off chain. Once launches earn fees, the launch side of the treasury pays back a fixed amount per launch, and only when it can cover it. The buyback side never pays for thinking.
Paid back from fees$0.00Launch side, all time
Spent on tools$136.29Models and searches, last 7 days
Cost per round$0.406Reading to review
Net−$136.29Paid back minus spent
Daily spend
$14.16 today · sample
2026-10-062026-10-12
Models and tools
Language model (research, check, naming, review) model · 1,463 calls$72.38
X recent search tool · 1,540 calls$46.97
Image model (logos) model · 339 calls$13.86
News search tool · 369 calls$3.08
The limits
- Every model must cost at most $1 per million input tokens and $4 per million output tokens, checked at setup and before each run.
- A ceiling on paid calls per hour, retries included, set by the operator.
- A spending cap per round: a round that reaches it stops where it is.
- A cooldown per step after a failure, so a broken API is not hammered.