Macropay nets your inference and infrastructure cost against revenue on the same ledger that bills it — so you see gross margin per agent, per model, per customer in real time. Spot the freeloader on Claude Opus before they cost you payroll.
Billing tools know your revenue. Observability tools know your cost. Macropay is the only ledger that holds both on one primitive — so the subtraction is real, not a spreadsheet guess.
Attach provider cost to every metered event — model token prices, GPU-minute rates, third-party tool fees. Macropay carries it alongside the revenue line for that same event.
await macropay.cost({ event: "gpt-4o.out", unit: "1K_tokens", provider: "openai", amount: 0.0100 // your cost});
Group margin by agent, model, customer, plan or cohort. Filter to a single account to see exactly who's profitable and who's burning your inference budget.
Teams cut their worst-margin cohorts and reprice freeloaders within a week of switching on insights.
Margin telemetry turns gut-feel pricing into evidence.
See the real cost of a plan before you ship it, and reprice tiers that quietly lose money.
Attribute inference spend to the agent or feature driving it — and decide what's worth keeping.
Gross margin, LTV and CAC on live billing data, not a month-old export.
Flag accounts whose usage outruns their plan before renewal, and act on it.
Blended gross margin everyone agrees on, because it's computed where the money moves.
Show per-agent margin and payback straight from the ledger in your next update.
Flat 4.5% + $0.50, all-in. Billing and insights on one ledger from day one.