The complete billing, metering, and payment operating system built for AI agents, API products, and usage-based pricing.
// Meter usage, bill customers, settle payments
const client = AgentPay.connect('sk_live_...)
await client.usage.ingest({
meter: 'gpt-4-tokens',
customer: 'cust_01H...',
value: 1847,
})THE PROBLEM
Legacy billing infrastructure was not designed for usage-based, agent-driven, real-time revenue models.
Traditional billing tools were built for seat-based SaaS. AI products need real-time metering, token-level granularity, and dynamic pricing.
When AI agents make purchasing decisions, you need approval flows, budget controls, and spend reservations that traditional payment rails cannot provide.
Billing, metering, approvals, settlements, and dunning scatter across 5+ tools. Each integration is a liability.
THE SOLUTION
Metering, pricing, billing, payments, and agent spend controls in a single operating system.
Ingest millions of events. Aggregate by meter, customer, and time window.
Tiered, volume, per-unit, formula-based. Price books with customer-specific overrides.
Budget-capped wallets for AI agents. Reserve-before-spend. Approval workflows.
Automated invoicing, payment collection, escrow holds, refunds, and disputes.
Full tenant isolation with test/live environment separation across all resources.
REST APIs, webhooks, API keys, domain events. Everything is programmable.
ARCHITECTURE
A composable architecture where each layer is independently deployable and horizontally scalable.
FOR DEVELOPERS
RESTful endpoints, typed SDKs, real-time webhooks, and a domain event system that makes every state change observable.
| Endpoint | Method | Description |
|---|---|---|
| /v1/products | GET | List products |
| /v1/products | POST | Create product |
| /v1/meters | GET | List meters |
| /v1/usage-events | POST | Ingest usage |
| /v1/invoices | GET | List invoices |
Start with the console to explore products, meters, agents, and billing flows.