Netlify’s free plan is a hard-capped tier: when the monthly allowance runs out, the site pauses rather than generating a bill - which is the safest possible design for a free tier, and the thing you need to know before you put anything important on one.
Free hosting tiers are one of the better deals in software, and they are also where a surprising number of production incidents begin. Not because the tiers are dishonest, but because people read the marketing page and skip the limits page. The useful skill is not knowing this one tier - it is knowing how to read any of them.
Table of contents
- The two kinds of free tier, and why the difference matters more than the numbers
- What actually gets metered
- The four questions to ask about any free tier
- Where a free tier is genuinely the right call
- How this fits the rest of the stack
- FAQ
The two kinds of free tier, and why the difference matters more than the numbers
Every free hosting tier is one of two designs, and they fail in opposite directions.
- Hard-capped. You get an allowance. When it is gone, the service stops until the next billing period or until you upgrade. You can never receive an unexpected bill. You can absolutely receive an unexpected outage.
- Soft-capped with overage. You get an allowance, and past it you are billed per unit. The site stays up. The bill does not stay predictable.
Netlify’s free plan is the first kind, deliberately - the allowance is a hard limit that cannot be exceeded or incur cost. That is the right default for a free tier, and it is why the free plan is a poor fit for anything with a deadline attached. A paused site at the wrong moment is worse than a small invoice.
The soft-capped design is where the horror stories come from. There is a well-known thread of a developer receiving a six-figure invoice for a static site after a traffic spike. That is an extreme case involving a scraper and a plan with overage enabled, but the mechanism is ordinary: metered bandwidth, no ceiling, no alert until the number was large.
What actually gets metered
Free tiers are rarely limited by one number. They are limited by whichever dimension you hit first, and it is usually not the one on the marketing page. Modern platforms - Netlify included - have moved toward a single unit of account that several activities draw from.
- Bandwidth. Data served to visitors. The dimension people expect, and often the one they underestimate because images and video dominate it.
- Build minutes or deploy count. Every push that triggers a build consumes allowance. A chatty CI setup burns this faster than traffic does.
- Function or compute time. Server-rendered pages and API routes are metered by execution time, not by request. A slow function is expensive twice.
- Requests. Page views, API calls, redirects, and every asset request. High-traffic sites with many small assets hit this before they hit bandwidth.
- Team members and concurrent builds. Not usage at all, but the limit that most often forces an upgrade once a second person joins.
The trap is assuming a free tier is limited by traffic. Often it is limited by seats or by build frequency, and you discover that on the day you hire someone.
The four questions to ask about any free tier
- What happens when I exceed it? Pause or bill. This is the single most important property and it is usually one sentence deep in an FAQ.
- Which dimension will I hit first? Estimate your real bandwidth, build frequency, and request volume. The binding constraint is rarely the advertised one.
- Can I set an alert or a ceiling? If overage exists, find out whether spending caps and notifications are available, and turn them on before you need them.
- What does the first paid step cost? The relevant number is not the free tier, it is the plan you land on when you outgrow it. Some ladders step gently and some step hard.
Do this once, write the answers down, and the free tier becomes a tool rather than a gamble.
Where a free tier is genuinely the right call
None of this is an argument against free hosting. For a personal site, a portfolio, documentation, a side project, a demo for a client, or an open-source project’s landing page, a hard-capped free tier is close to ideal. The site is static, the traffic is modest, and a pause would be an annoyance rather than an incident.
It is also the right call for evaluation. Deploy the real project, look at the build logs, see how the routing behaves, and find out whether the platform’s rewrite semantics match what your framework expects. That is worth far more than reading comparison articles.
Where it stops being the right call is the moment someone else depends on the site being up. A client’s marketing site, a signup page behind an ad spend, an API a mobile app calls. At that point you want a plan with a predictable price and no pause behaviour, and you want to know that number before the traffic arrives rather than after.
How this fits the rest of the stack
The useful exercise is not comparing free tiers, it is finding out what the thing costs at the size you expect to be in six months. The RunxBuild hosting calculator puts the service, the database, the storage, and the bandwidth on one page so that number is concrete rather than a range. On RunxBuild, static sites include 120GB of bandwidth and are $0.10/GB after that, and the general plan ladder starts with 15 free days each month before a $1.80 charge, so there is a free step without a cliff behind it.
Useful related references:
- Static Variables in Java: Shared State You Did Not Plan For
- PostgreSQL 13 End of Life: Support Ended, So Plan the Upgrade
- Domain Migration: DNS Plan, 301 Redirects, and SEO Preservation
- Services on RunxBuild
FAQ
Is the Netlify free plan really free?
Yes, and it is hard-capped by design: the monthly allowance cannot be exceeded or incur charges. When it runs out, sites pause until the next month or until you upgrade. That protects you from surprise bills and exposes you to surprise downtime, which is the trade every hard-capped tier makes.
Can you use a free hosting tier for a commercial site?
Netlify’s free plan permits commercial projects, and plenty of small business sites run happily on free tiers. The question is not permission but risk: if a pause during a busy week would cost you real money, the free tier is the wrong tool regardless of what the terms allow.
What usually causes people to exceed a free tier?
Rarely the thing they expected. Large unoptimised images pushing bandwidth, a CI setup that triggers a build on every commit to every branch, server-rendered routes consuming compute time, or simply adding a second team member. Check seats and build frequency, not just traffic.
How do I avoid a surprise hosting bill?
Prefer hard-capped plans for anything experimental. Where overage exists, enable spending limits and billing alerts before you need them, and check whether auto-recharge is on by default. Then estimate bandwidth honestly - it is the line item people underestimate most.
When should I upgrade from a free plan?
When someone other than you depends on the site being up, when a second person needs to deploy, or when you are within roughly half of any limit. Upgrading before you hit a cap is a scheduling decision; upgrading after you hit one is an incident.