One endpoint for every coding agent you run

Swap one environment variable and your existing agents keep working — the tools stay the same, the billing changes. Usage is metered per token and drawn from a single prepaid balance instead of a stack of monthly plans.

Every request is attributable: which model answered, how many tokens it cost, and what the running total is, visible as it happens rather than at the end of a billing cycle.

Requests route automatically across upstream capacity, so a provider hiccup does not end a session.