How It Works

Four steps to budget enforcement

From payment provider connection to live API verification in minutes.

Enforcement: 402 Hard Stop

Verification: <50ms

Budget Model: Dollars

Connect your payment provider

One-click OAuth with Polar. Subscriptions and plans sync automatically.

Automatic webhook registration
Sync existing subscriptions
Stripe & LemonSqueezy in beta

Create API keys with budgets

Issue keys with dollar-based budgets, scopes, and environment separation.

Prefixed keys (hk_live_, hk_test_)
Dollar-based budget limits
Per-environment isolation

Verify requests in < 50ms

Call POST /v1/verify from your middleware. Get validity, remaining budget, and entitlements.

Sub-50ms response times
remaining_budget_usd for enforcement
Entitlements from subscription plans

Enforce with 402

Return 402 Payment Required when budget is exceeded. No more runaway costs.

Hard stop at budget limit
Audit trail for all decisions
Usage tracked automatically

Integration

Add to your middleware

A few lines of code to verify every request and enforce budgets.

middleware.ts

const result = await holdify.verify({
  key: apiKey,
  resource: 'chat.completions',
});

if (!result.valid) {
  return res.status(401).json({ error: 'Invalid key' });
}

if (result.remaining_budget_usd <= 0) {
  return res.status(402).json({
    error: 'budget_exceeded',
    message: 'Payment Required',
    resetAt: result.reset_at
  });
}

// Budget verified - proceed with request
const completion = await openai.chat.completions.create({
  model: 'gpt-4',
  messages: req.body.messages
});

Ready to protect your AI revenue?

Start free with 150K tokens/month. No credit card required. Setup in under 5 minutes.

Get Started Free Read the Docs