Move the sliders below: users, actions each, AI calls per action. Watch the monthly bill.

Model

Hosting + database

fine for a demo; pauses or throttles when real traffic arrives

Estimated monthly bill$0.21

  • AI calls (500)$0.21
  • Hosting$0
  • Database$0

AI calls are 100% of the bill and they scale with users; hosting is flat until you outgrow a tier. Cheapest lever: fewer calls per action, then a smaller model for the easy steps.

Round teaching tiers, not live prices (assumes ~1,200 input + 400 output tokens per AI call). The lesson links the real pricing pages; this shows the shape of the bill.

Hosting is renting the always-on computer your app runs on. The good news: most of it is free until people actually show up.

The calculator above uses round teaching tiers (a roughly $20/month host tier, a roughly $25/month database tier) to show the shape of a bill, not a quote. Real prices move, so check the live pages: Cloudflare Workers, Vercel and Supabase.

Free tiers really are generous. Supabase's free plan (checked 25 Sep 2026) includes a 500 MB database and 50,000 monthly active users, and Cloudflare's pricing docs put the paid Workers plan at "a minimum charge of $5 USD per month for an account." A small app with real users can cost nothing for a long time.

What you actually pay for

Compute / hosting

Roughly when it starts costing
Free tier, then a few dollars a month

Database

Roughly when it starts costing
Free tier, then ~$25/mo at real scale

Domain name

Roughly when it starts costing
~$10–15 a year, always

AI model calls

Roughly when it starts costing
Per token, from the first call, and it grows with users

The line that surprises people is AI. Every model call is billed by the token, for what you send and what comes back, so a feature that calls a model on every action costs more as more people use it. Hosting is a flat-ish rent; AI is a per-use meter.

That's why the calculator's biggest mover is usually "AI calls per action", not "users". Two apps with the same traffic can have very different bills depending on how often each one calls a model.

Try it with the rules off. In the sandbox, set AI calls to zero and raise users to ten thousand; the bill barely moves. Then set users low and AI calls high; watch it climb. That's the real cost driver.

Check yourself

0/4 got it

Saved on this device only. No account, no streaks.

Next in making it real: Environment variables and secrets.