Usage Metering for AI
Your AI feature shipped.
Your billing didn't.
Track usage per customer. Enforce limits. Bill with confidence.
Two API calls — and you never build metering infrastructure again.
api/chat.ts
// Check limits before the call
const check = await limitry.limits.check({
customerId: "cust_acme",
meterId: "llm-tokens",
});
if (check.allowed) {
const completion = await openai.chat.completions.create({ ... });
// Record what happened
await limitry.events.record({
customerId: "cust_acme",
eventType: "llm.completion",
value: completion.usage.total_tokens,
});
}