Cloud Uptime Check Pricing on AWS, Azure and Google Cloud

The same uptime check costs between $2.75 and $157.56 a month on the three big clouds, and AWS is at both ends of that range. One HTTPS URL checked every minute from three locations is $2.75 as a Route 53 health check outside AWS, $39.42 on Google Cloud Monitoring once its free million executions are used, and $157.56 as three CloudWatch Synthetics canaries. Azure Application Insights cannot check more often than every five minutes, where it costs $13.14.

Every figure below is a US list price in US dollars, read from each provider's own price list or pricing page, and every monthly figure uses the 730 hour month that AWS, Azure and Google all use in their own calculators. Where a provider adds charges in another service (Lambda for canaries, Log Analytics for Azure), we say so rather than guess the amount.

How the four cloud uptime checks are billed

The three clouds sell four different products for the same job, and only one of them is priced per check. The other three are priced per execution, which means the interval and the number of locations multiply straight into the bill.

Per execution, Google is the cheapest at $0.0003, Azure next at $0.0005, and a CloudWatch canary four times Google's rate at $0.0012. Route 53 does not have an execution price at all. Its checkers in several AWS Regions hit your endpoint about every two seconds between them, and you pay the same flat monthly fee whatever that adds up to.

What one uptime check costs on AWS, Azure and Google Cloud

Take one public HTTPS website, hosted outside the cloud you are pricing, checked from three locations. That is the minimum Google allows and a common choice everywhere else.

The arithmetic is short enough to check by hand. A 730 hour month has 43,800 minutes. At one-minute checks from three locations that is 131,400 executions, which is $39.42 at Google's rate and $157.56 at the canary rate after its 100 free runs. At five minutes it is 26,280 executions: $7.88 at Google, $13.14 at Azure and $31.42 for canaries. Route 53 ignores all of it and charges $0.75 for the check plus $2.00 for HTTPS.

So the honest answer to "which cloud is cheapest for uptime checks" is Route 53, by a wide margin, as long as all you need is pass or fail on an HTTPS URL. The full option ladder, including what string matching and the fast interval add, is in our Route 53 health checks pricing breakdown.

Is Google Cloud uptime monitoring free?

For slow checks, effectively yes. Google gives each project 1 million executions a month, and that is the only free allowance on this list large enough to run a real fleet on. How far it goes depends on the interval:

• Every 15 minutes from 3 regions: all 100 checks a project allows, 876,000 executions, stay free.
• Every 5 minutes from 3 regions: 38 checks fit.
• Every minute from 3 regions: 7 checks fit.
• Every minute from Global (6 regions, the console default): 3 checks fit.

Past the free million each extra one-minute check costs $39.42 a month from three regions or $78.84 from Global. That cliff is the part most teams miss, because the first several checks cost nothing and the bill only appears when the allowance runs out. Our Google Cloud Monitoring pricing calculator for GCP uptime checks shows exactly where it lands for your interval and region count.

Azure has no equivalent allowance at all: the first execution of the first test is billed. AWS gives 100 canary runs, which at a one-minute interval lasts 100 minutes. Route 53's free 50 checks apply only to endpoints running inside your own AWS account, never to a site hosted elsewhere, and never to the HTTPS option.

Cloud uptime check cost for 20 websites

Twenty URLs is where most small SaaS and agency teams land once they count the marketing site, the app, the API, the status of a few customer domains and the login page. Here is that fleet with alerting included where the cloud bills it.

Two things stand out. Google at five-minute checks costs nothing for this fleet, which is a genuinely good deal if five minutes of blind time is acceptable. And the moment you ask Google for one-minute detection the same fleet costs $488.40 a month, twenty times what a flat per-monitor plan charges for the same interval from twice as many regions. The Route 53 row is cheap for a different reason: its alarms are $0.10 each after ten free, which is the whole of what alerting costs on AWS, as we priced out in CloudWatch alarm pricing for downtime alerts.

Per-execution meters also make the bill hard to predict, since anyone with console access can change an interval or add a region and the cost moves with it. If your cloud bills already surprise you, it is worth having real-time alerts when a metered bill starts climbing, separate from the monitoring itself.

What each cloud check actually verifies

Price per check only makes sense next to what the check does. They are not equivalent.

Route 53 is cheapest because it was built to steer DNS, not to tell people a site is down. It is the only one here that can pull a failing record out of rotation, and the only one that keeps passing on an expired certificate. Google and Azure both check certificates and content. Canaries can check anything, because they run your script in a Lambda function, which is also why they cost the most.

Which cloud uptime check should you buy?

Match the product to the job rather than to the cloud you already use:

• Route 53 when the check drives DNS failover or weighted routing. Keep the option list short, since every option is $2.00 a month per check outside AWS.
• Google Cloud when your services run on GCP and five or fifteen minute detection is enough. Inside the free million it is the best value here.
• Azure Application Insights when the app is already instrumented with it and the test results belong next to its telemetry. Budget $0.0005 per execution from the first one; Microsoft's recommended five locations at five minutes is $21.90 a month per URL, as the Azure Application Insights pricing breakdown works through.
• CloudWatch Synthetics for scripted journeys such as a login or a checkout, not for plain uptime. The CloudWatch Synthetics pricing post covers the Lambda and storage charges that ride on every run.

What none of the four gives you is a customer-facing status page, alerts that reach Slack or SMS without wiring a second service, or a monitor that lives outside the account it is watching. If the account has a billing hold or someone deletes the wrong role, cloud-native checks go quiet at exactly the wrong moment.

Is a cloud-native uptime check cheaper than an uptime monitoring tool?

Below about eight one-minute checks, often yes. Route 53 stays under $24 a month for up to eight outside HTTPS checks, and Google's free million covers seven one-minute checks from three regions. Past that point a flat plan is cheaper for every cloud except Route 53, and Route 53 is cheaper only because it skips certificate checks, content checks and a status page.

UptimeHub Starter covers 20 monitors checked every minute from six regions for $24 a month billed yearly or $49 monthly, and Pro covers 100 monitors every 30 seconds for $49 yearly or $99 monthly. Both include SSL and domain expiry warnings, alerts by email, Slack and webhook, and a branded status page, with no execution meter to watch. The full plan list is on the UptimeHub pricing page, and the wider market of per-run vendors is compared in the synthetic monitoring pricing comparison.

Where these numbers come from

Route 53, CloudWatch alarm and CloudWatch Synthetics rates come from the AWS Price List for US East (N. Virginia). The Google Cloud rate, free allowance and postponed alerting charges come from Google Cloud's Observability pricing page, and the counting rule from Google's own pricing examples, which our model reproduces exactly ($94.20 for a 10 member group checked every minute from the USA location). The Azure rate comes from the Azure retail price list for East US. We do not estimate the Lambda, S3 and CloudWatch Logs charges that canaries add, Azure data ingestion, or any negotiated discount.