Predictable hosting costs
Hosting bills surprise people for a few recurring reasons: metered usage, traffic nobody invited, and limits that behave differently from what their name suggests. None of this is universal — it depends on the provider and the plan.
Two ways to bill
By size: you pay for a server or container of a given size, whatever the traffic. The bill is stable; the risk is running out of capacity.
By use: you pay for what is consumed above an included quota — data transfer, requests, function time, build minutes, database compute. The bill follows traffic, including traffic from bots and crawlers. One provider states that requests served before an attack is recognised, and bot traffic, are billed (example).
What a spend cap really does
Where a hard cap exists, reaching it usually takes the site down: production can be paused until you resume it by hand (example), every site can be paused when credits run out (example), or every workload can go offline (example).
Elsewhere, budgets are alerts and do not stop spending (example, example), some providers have no cap at all (example), and some alerts arrive hours after the spend (example). A cap on one item may leave others outside it (example).
So the real choice is between a bill that can grow and a site that can stop. Know which one your plan makes for you.
Credits that pay for more than building
Questions to ask before you choose
- What is metered, and what is included each month?
- What happens at the limit — a higher bill, a slower app, or a paused site?
- Is the cap checked in real time, or can spending pass it before it applies?
- Does attack or bot traffic count toward usage?
- Are any paid options switched on by default for new projects (example)?
- Which costs come from other providers your app calls — AI APIs, payments, e-mail — and are billed by them?
Next step
HostSpry bills hosting per project, not per visitor. Tell us about your app for a migration review.