Budgets and spend

Agents spend real money, so spend is capped twice: a budget on every run, and trailing limits on totals. Both ship with defaults you can tighten in Settings → Spend or per agent in its config.

How budgets work

Every run resolves a budget before it starts: the config's limits.run if set, otherwise your account default ($50, also the platform maximum per run). As the agent works, each model call debits the budget in real time. A run that reaches its cap stops immediately and records budget_hit; the run timeline shows which step spent what.

limits:
  run: 5.00   # this run may spend at most $5

For a weekly-digest agent, $1 is usually plenty. For an agent that opens PRs, budget for test runs too.

To give a single run more room without editing the config, pass config_override_yaml when starting or replaying it.

Trailing spend limits

Account-wide caps on total spend over rolling windows, checked before every run. A run that would exceed one is cancelled before a sandbox exists.

Adjust them in Settings → Spend, up to a platform ceiling of $1,000 / $10,000 / $40,000.

Per-agent trailing limits

A config caps its own trailing spend with limits.day, limits.week, and limits.month (the same 1/7/28-day windows, same ceilings). This is the difference between "the account may spend $250 today" and "this one agent may spend $2 this week":

limits:
  run: 1.00
  day: 2.00
  week: 10.00
  month: 30.00

Omit any value to inherit the default: the account default for run, the platform ceiling for the windows.

Reading spend

Each run stores its cost broken down into token spend, sandbox CPU, sandbox memory, and Ellipsis fee (see How a run executes). agent budget and GET /v1/budget show current spend against every limit.