QuotaMint

we bill per call minute and the case that worries me isn't the deduction logic itself, it's what happens when the usage event arrives twice, a webhook retry, a flaky network on our end, whatever, and both attempts hit "atomic credit deduction" with the same idempotency key. does the second call return the original result without re-deducting, or does it just fail loudly? for us a silent double-charge would be way worse than an error we can catch and retry

the part i ended up rebuilding twice wasn't the deduction, it was stacking. on our job board employers get one free test listing, and later we added a promo code worth a few free listings. first version just added them up, so someone with the test post plus the code got one more than we meant, and the code's credits never expired because nobody had decided whether the clock starts when the code is issued or when it's redeemed. ended up capping free listings across all sources and expiring promo credit 30 days after activation, which meant tracking every grant separately instead of one balance. does quotamint keep grants as separate buckets with their own expiry, and if so which one gets consumed first when a user has both?

I built QuotaMint because I kept seeing the same thing: every SaaS or AI product ends up rebuilding credits, usage limits, plan access, and usage tracking in slightly different ways.
The goal was to make that part simple. QuotaMint gives you one API to handle credits, limits, plans, and feature access, while you keep using your existing billing provider like Stripe, Paddle, or Dodo.
I wanted it to feel lightweight, easy to integrate, and useful from day one, especially for small teams and solo founders.
Would love to hear what you think, especially if you’re building a SaaS, AI app, or API product