Meter customer usage, launch PAYG billing, reduce token spend, forecast spend, catch anomalies, and give customers drop-in usage dashboards.
No credit card required. Then $0.001 per call.
Track API and app usage per customer, detect new traffic patterns, and forecast when accounts may run out before period end. Alert teams and customers via Slack or email so you can upgrade active accounts before limits turn into churn.
Compress prompts and route standard work to faster, lower-cost models while reserving premium usage for harder tasks. Track saved tokens and cost impact without storing raw prompts.
Set quotas, rate limits, and block or throttle overages before traffic reaches your vendor bill. Let successful trials grow without one account burning margin or abusing your API.
Server-side integration
Start a usage check, enforce the customer's plan, optionally compress the prompt, make the vendor call, then report usage and cost. Use the SDK or plain HTTP in trusted server runtimes.
Supports LLM calls, tokens, model tiers, compression savings, rate limits, API calls, searches, audio/video minutes, and custom meter types.
AI prompt kit / reference →Begin the call, optionally compress locally, invoke your model, then report usage.
const begin = await usageTap.beginCall({
customerId,
feature: "chat",
requested: { premium: true },
});
if (!begin.data.allowed.premium) {
throw new Error("Premium calls exhausted for this customer");
}
const compressed = await usageTap.promptCompress({
callId: begin.data.callId,
input,
});
const model = "openai/<model>";
const response = await openai.responses.create({
model,
input: compressed.compressedInput,
});
// Model name is illustrative; UsageTap supports multiple providers via your configured AI stack.
await usageTap.endCall({
callId: begin.data.callId,
modelUsed: model,
inputTokens: response.usage?.input_tokens ?? 0,
responseTokens: response.usage?.output_tokens ?? 0,
});
Every AI feature costs real money per call. UsageTap lets you offer safe free trials, charge for real usage, and introduce committed plans when an account's behavior becomes predictable—all driven by policy, not code changes.
01
Give prospects room to test while quotas and rate limits protect your margin.
02
Charge for real usage after allowance, whether it is API calls, tokens, or app actions.
03
Forecasts show which accounts are likely to exceed limits or need a larger package.
04
Offer included allowance plus PAYG overage so heavy accounts get predictability.
UsageTap packages the metering, guardrails, customer UI, alerts, and billing sync teams usually build piecemeal.
| Capability | DIY | UsageTap |
|---|---|---|
| Usage metering | Custom event pipeline & aggregation | Server-side SDK or HTTP |
| Plan & tier logic | Entitlement service and upgrade flows | Dashboard-managed policies |
| Stripe integration | Webhooks & reconciliation | Automatic sync |
| Customer usage UI | Build from scratch | Drop-in React widgets |
| Prompt compression & model tiers | Custom heuristics & savings math | Opt-in SDK helper + dashboard savings |
| Rate limiting & overages | Custom middleware | Policy-driven controls |
| Forecasts & anomaly alerts | Statistical pipeline and notification logic | Built-in Slack and email alerts |
| Timeline | 3–6 months | Working integration in days |
Customer usage signals
UsageTap watches expected usage patterns by customer. Send Slack or email alerts when traffic changes suddenly or a customer is forecast to exhaust allowance before renewal, so your team can offer a plan change and customers can adapt early.
Spot spikes, drops, or new workloads that need support, pricing, or abuse review.
Warn customers and account teams before allowance is exhausted, with CTAs to upgrade or top up.
Plan- or customer-level caps per call, token, and capability—no custom code per tier.
Block, throttle, or bill overage. Switch policies without engineering sprints.
Drop‑in React components for usage vs allowance, forecast charts, and upgrade prompts.
Compress prompts and separate standard from premium usage for lower-cost model routing.
Forecast future spend in dollars and API calls—act before overages become churn.
Detect usage diverging from forecasted expectations and notify teams before surprises spread.
Alert internal teams and customers when patterns change or forecasts show allowance risk.
Stream calls, tokens, and cost to any OTLP endpoint—Datadog, Grafana, New Relic.
Free tool
Upload your OpenAI or Anthropic billing CSV and get an instant forecast with anomaly checks before you integrate UsageTap. No signup required.
As your customers grow
AI-powered products don't behave like traditional software. One customer may send a few prompts while another runs thousands of calls overnight. Adaptive Pricing lets you start with PAYG and introduce commitment only when usage becomes predictable.
Let customers begin with free + PAYG so they only pay when they get value. No lock-in to one monetization path.
Move stable accounts into committed plans with included allowance and PAYG overage. No forced blanket upgrades.
Committed tiers work as a savings option. Customers who prefer flexible PAYG stay there.
Automatic, manual, or no migration at all. You choose the policy that fits your GTM.
If you are launching an AI app or API, UsageTap gives you the metering, controls, and customer-facing usage layer to move from free trials to paid usage with less margin risk. Shape the product with us.
.webp&w=128&q=75){kind=small}
Founder University