AAI Usage ControlUsage visibility and runtime limits
Sign inCreate account
For AI products with real usage costs

See who is consuming your AI budget and control what happens next.

Track usage across free, trial, paid, and enterprise customers. Set budgets and quotas by account, plan, feature, or application. Enforce limits before a small group of users destroys your margins.

Create demo account
✓ Works for subscribed and non-subscribed users✓ No proxy required to start✓ Add runtime checks when ready
AI budget status
$13,988 of $20,000
70% used
Requests2.4M
Blocked18.2K
Saved by controls$3,742
Free plan monthly budget

Block after $1.50 per user

Enabled
UserTypeSpendAction
usr_7J92Free$3.21Blocked
acme_teamPaid$172Throttled
trial_128Trial$0.88Allowed
Provider invoices show total spend.They do not tell you which customer caused it or stop the next expensive request.
One platform

Observe usage, understand cost, and enforce policy.

Start with lightweight tracking. Add real-time checks only where you need control.

◎
Usage visibility
Attribute requests, tokens, retries, failures, and spend to customers, plans, features, and applications.
▤
Budgets and quotas
Define monthly spend, request, token, or feature limits for free users, trials, paid plans, teams, or individual accounts.
↯
Runtime decisions
Check whether the next request should be allowed, blocked, throttled, or capped before you send it to a model provider.
◇
Revenue context
Connect subscription systems and compare AI usage with plan revenue so expensive accounts become obvious quickly.
Two SDK paths

Adopt at your own pace.

01

Track only
Send usage events after each model attempt. Get cost attribution and dashboards without changing request flow.
await control.track({
  userId,
  feature: "research",
  plan: "free",
  model: "gpt-4.1-mini",
  inputTokens,
  outputTokens
});

02

Check and enforce
Call the decision endpoint before the model request. Your app receives a deterministic policy decision in milliseconds.
const decision = await control.check({
  userId,
  accountId,
  feature: "research"
});

if (!decision.allowed) {
  return showUpgradePrompt();
}
Policy examples

Control AI usage at the level that matches your business.

Free users

$1.50 monthly AI budget
Block requests after a user reaches the limit.

Trial accounts

50 research requests
Allow core usage while protecting expensive features.

Paid plan

$20 per user, per month
Throttle unusually heavy accounts instead of blocking everyone.

Enterprise team

$1,000 shared budget
Let admins distribute usage across teams and members.
Connect your stack

Usage, billing, and identity stay in sync.

SStripe
RCRevenueCat
◉OpenAI
AAnthropic
GGemini
APICustom events
Interactive sandbox

Explore policies, usage data, and runtime decisions with a populated demo.

Create an application, inspect heavy users, simulate policy checks, and see how limits affect free, trial, and paid accounts.

Create account