Synthetic monitoring tools: synthetic transaction monitoring and the best synthetic monitoring software, priced per check

Short answer: almost every synthetic monitoring tool bills per check execution, and the rate decides your bill far more than the plan you pick. One HTTP endpoint checked every 30 seconds for a month costs $8.77 on Splunk, $15.78 on Checkly, $43.83 on Datadog or Grafana Cloud, $87.67 on Dynatrace and $438.33 on New Relic. The same $59 buys 100 monitors on a flat plan. Metered tools are the right answer when you run a handful of scripted browser journeys. They get expensive fast when what you actually need is broad availability coverage.

Check a URL first ▸

Last updated September 2026 · every rate below read from the vendor's own published price list

alertping ▸ run check

live

▸ type a domain and run a real-feel check

▸ probes from 3 regions · FRA · IAD · SIN

▸ waiting…▌

If ever goes down, you get:

Alert fired ▸ 2 channels · 6.2 s after first failure

AlertPing app

● DOWN : HTTP timeout confirmed from 3/3 regions (FRA, IAD, SIN). Incident opened.

sms · on-call

AlertPing: DOWN. Confirmed 3/3 regions . First fail: Frankfurt.

the number that decides your bill

What synthetic monitoring tools charge per check execution

Vendors publish plan prices because plan prices are comparable. The meter is where the money goes, and the meters differ by a factor of fifty across tools that look similar in a feature grid. Every rate here is the US list price on the vendor's own pricing page, converted to a cost per single execution so the columns can actually be compared.

Tool	Published rate	Per HTTP/API check	Per browser/scripted check	Billing unit
Splunk Observability	$0.0001 per uptime request	$0.0001	$0.0045 per browser action	Per request, per action
Checkly (prepaid)	$1.80 / 10k API, $4.00 / 1k browser	$0.00018	$0.004	Per check run
Checkly (automatic overage)	$2.60 / 10k API, $6.50 / 1k browser	$0.00026	$0.0065	Per check run
Datadog Synthetics	$5.00 / 10k API, $12.00 / 1k browser	$0.0005	$0.012	Per test run
Grafana Cloud Synthetic Monitoring	$5.00 / 10k API, $50.00 / 10k browser	$0.0005	$0.005	Per execution, per probe location
Dynatrace	$0.001 per synthetic request	$0.001	$0.0045 per browser action	Per request, per action
New Relic	Ping monitors free, $0.005 per non-ping check	$0.005	$0.005	Per check beyond the allowance
AlertPing	Flat plan, from $19/mo	No meter	No scripted browser checks	Per monitor, per plan

Two things in that table catch people out. Grafana Cloud bills per probe location, so a check you think of as one monitor is three or five line items on the invoice. And Checkly charges two different rates for the same execution depending on whether you bought the capacity in advance: browser runs are $4.00 per thousand prepaid and $6.50 per thousand when they land as automatic overage, a 62.5% penalty for not forecasting.

the comparison nobody prints

How many 30-second HTTP monitors $59 a month buys

A check running every 30 seconds fires 87,666 times in a 30.44-day month. Multiply that by each vendor's per-execution rate and you get the honest answer to the question buyers actually have, which is not "what does the plan cost" but "how much of my estate can I actually watch for this budget".

Tool	Cost of one 30-second monitor	Monitors $59 a month buys
AlertPing Team	Included, no meter	100
Splunk Observability	$8.77	6.7
Checkly (prepaid API runs)	$15.78	3.7
Checkly (automatic overage)	$22.79	2.6
Datadog Synthetics	$43.83	1.3
Grafana Cloud (one probe)	$43.83	1.3
Dynatrace	$87.67	0.7
New Relic (non-ping check)	$438.33	0.1

where they win

New Relic wins outright on plain ping

New Relic ping monitors are free and unlimited on every edition, which beats Datadog, Dynatrace, Checkly and us for that one job. The $0.005 rate in the table applies only to non-ping checks. If a plain reachability ping is genuinely all you need and you already run New Relic, stop reading and use it.

where they win

Splunk undercuts a flat plan below 6.7 monitors

At $0.0001 per uptime request, Splunk's meter is cheap enough that a small estate costs less than our $59 Team plan. Six 30-second monitors are $52.60 a month there. The crossover is 6.7 monitors, and above it the flat plan pulls away permanently.

where meters hurt

Interval is a price, not a setting

On a metered tool, halving the interval doubles the bill, so budget quietly pushes teams onto five-minute checks. That matters: one missed five-minute check is 5 minutes of unrecorded downtime, which is 114% of the entire 4m 23s monthly budget of a 99.99% target. The cheap interval cannot evidence the SLA you promised.

know what you are buying

Three kinds of synthetic check, three completely different prices

"Synthetic monitoring" covers three jobs that vendors price between 40 and 120 times apart. Buying the wrong one is the most common way teams overspend on this category, and it usually happens because a scripted browser test felt like the thorough choice for a job an HTTP check would have done.

cheapest

Uptime and protocol checks

HTTPS, TCP, DNS, ICMP, SSL and port checks. One request, one verdict. This is what answers "is it up", it is what your SLA report is built from, and it is the only class cheap enough to run on everything you own at a short interval.

Typical cost: $0.0001 to $0.001 per run

middle

API and multistep checks

An HTTP request with assertions on status, headers, response body and latency, optionally chained so one call feeds the next. This catches the failure where the page renders fine but the API behind it returns an empty payload, which a plain ping never sees.

Typical cost: $0.00018 to $0.005 per run

most expensive

Browser and transaction checks

A real headless browser walks a journey: log in, search, add to cart, check out. This is synthetic transaction monitoring proper, and it is the only thing that proves the funnel still works. It is also 40 to 120 times the price of an HTTP check.

Typical cost: $0.004 to $0.012 per run

What a scripted browser journey costs at each interval

This table is why almost nobody runs a checkout script every minute, and why teams that need 30-second detection use a cheap HTTP check for detection and a slow browser check for correctness. One journey, one location, list rates.

Interval	Runs per month	Checkly prepaid	Grafana Cloud	Dynatrace (1 action)	Datadog
Every 1 minute	43,833	$175.33	$219.16	$197.25	$526.00
Every 5 minutes	8,767	$35.07	$43.83	$39.45	$105.20
Every 15 minutes	2,922	$11.69	$14.61	$13.15	$35.07
Every 30 minutes	1,461	$5.84	$7.31	$6.57	$17.53
Every 60 minutes	731	$2.92	$3.65	$3.29	$8.77

Grafana's figure multiplies by the number of probe locations, and Dynatrace's multiplies by the number of actions in the script, so a five-step journey from three regions is a different order of magnitude again. Read your own script length and location count into these numbers before you commit to an annual plan.

the honest shortlist

Which synthetic monitoring software fits which team

Nobody is best at everything here, and a shortlist that says otherwise is selling. This is where each tool genuinely wins, including the cases where we are the wrong answer.

Tool	Best for	Weakest at	Entry price
Checkly	Playwright shops that want their existing browser tests running as monitors, in code, in CI	Broad availability coverage: included run allowances are small and the overage rate is 44% to 63% above prepaid	$24/mo annual
Datadog Synthetics	Teams already on Datadog who want a failed check to open into traces and logs in one click	Price per browser run, the highest in this table at $0.012	Add-on to a Datadog contract
Grafana Cloud	k6 users and teams that want probes inside their own network alongside metrics and logs	Per-probe billing makes multi-region checks quietly multiply	Free tier, then $19/mo plus usage
Dynatrace	Enterprises that want failed journeys correlated automatically with root cause across the stack	Per-action browser billing punishes long scripts	Consumption on a Dynatrace platform contract
New Relic	Anyone who needs plain ping monitors and nothing more, because those are free and unlimited	$0.005 per non-ping check is the most expensive HTTP meter here	Free edition, then per-user
Splunk Observability	Small estates where the $0.0001 request rate beats any flat plan, with log correlation attached	Costs cross a flat plan above roughly 6.7 monitors and it is sold into a platform contract	Metered on an Observability Cloud contract
AlertPing	Teams that need many endpoints watched at a short interval on a bill they can predict a year out	No scripted browser journeys, no APM, no RUM, no free plan	$19/mo flat, 30-second checks from $59

straight answer

What AlertPing does, and what it deliberately does not

What we do

HTTPS, TCP, port, DNS-resolvable host and SSL certificate checks, as often as every 30 seconds on Team and every 15 seconds on Business.
Keyword and status-code assertions, so a page that returns 200 while showing an error is still caught.
Every failure confirmed from three independent regions before it counts as downtime, which is what stops the 3am false page.
SMS, email, Slack and webhook alerts on every plan, with no credit packs and no per-message billing.
Cron and scheduled-job heartbeats, SSL expiry warnings and public status pages included rather than sold as modules.
A flat price per plan with no per-check meter anywhere, so the interval is a setting rather than a budget decision.

What we do not do

No scripted browser journeys. We do not run Playwright or Selenium scripts, so we cannot prove a checkout flow completes. If that is your requirement, Checkly and Datadog are the honest recommendations.
No APM, no distributed tracing and no real user monitoring. We tell you the service is unreachable or wrong, not why the third database call was slow.
No installed agent and no infrastructure metrics. Everything is measured from outside your network.
No free plan. If you need zero-cost monitoring for a side project, UptimeRobot and New Relic ping monitors both beat us there and we would rather say so.

The pairing most teams land on is the cheap thing running constantly and the expensive thing running slowly: flat-priced HTTP and port checks every 30 seconds across the whole estate for detection and SLA evidence, plus a small number of scripted journeys every 15 or 30 minutes for correctness. That combination costs a fraction of running everything through a metered browser tool, and it detects faster. There is a longer version of that argument in our writeup on pairing uptime monitoring with an APM platform.

questions buyers actually ask

Synthetic monitoring tools: common questions

What is synthetic monitoring?

Synthetic monitoring runs scripted requests against your site or API from outside your infrastructure, on a schedule, whether or not real users are active. A probe sends the request, checks the response against rules you set, and alerts when the rules fail. It is active measurement, which is why it detects an outage at 4am on a Sunday when no real traffic exists to reveal one.

What is synthetic transaction monitoring?

Synthetic transaction monitoring drives a real headless browser through a multi-step user journey such as login, search, add to cart and checkout, then asserts that every step completed. It proves the funnel works rather than that the homepage loads. It is the most expensive class of synthetic check because each run costs browser compute, typically $0.004 to $0.012 per execution.

How does synthetic monitoring work?

A probe in a data center runs your check on a fixed interval, records the status code, latency and response body, and compares them against your assertions. If an assertion fails, most tools retry or confirm from another region before raising an incident, then notify by SMS, email, Slack or webhook. The result is a continuous availability record independent of your own servers.

How much does synthetic monitoring cost?

Metered tools charge $0.0001 to $0.005 per HTTP check and $0.004 to $0.012 per browser run, so one endpoint checked every minute for a month is roughly $4 to $220 depending on the vendor, and one browser journey at the same interval is $175 to $526. Flat-plan tools remove the meter entirely and price by monitor count instead, from $19 a month.

What are the best synthetic monitoring tools?

It depends on which of the three check types you actually need. Checkly leads for Playwright-based transaction monitoring, Datadog and Dynatrace lead when failures must correlate with traces, New Relic wins for free plain ping, Splunk has the cheapest HTTP meter, and flat-plan tools like AlertPing win once you need dozens of endpoints watched at a short interval.

What is the difference between synthetic monitoring and real user monitoring?

Synthetic monitoring is active and runs on a schedule from a fixed probe, so it works with zero traffic and gives a clean, comparable baseline. Real user monitoring is passive and reports what actual visitors experienced, including their devices and networks. RUM cannot detect an outage nobody is awake to hit, and synthetic cannot tell you the site is slow on a specific carrier. Most teams need synthetic first.

Do I need synthetic monitoring if I already have uptime monitoring?

Uptime monitoring is a form of synthetic monitoring, the cheapest one. You need the more expensive kinds only when a working URL is not sufficient proof: a checkout that silently fails at the payment step, an API that returns 200 with an empty array, a login that breaks after a session change. If those failure modes would cost you money, add scripted checks on top rather than replacing what you have.

How often should synthetic checks run?

Match the interval to the availability target you promised, because one missed check is the smallest outage you can record. A 99.99% target allows 4 minutes 23 seconds a month, so a five-minute interval cannot evidence it: a single failed check already exceeds the whole budget. Thirty seconds is the practical floor for SLA reporting, and browser journeys can run far slower because they are checking correctness rather than reachability.

Availability coverage without a per-check meter

AlertPing checks websites, APIs, ports and SSL certificates as often as every 30 seconds, confirms every failure from three regions before it counts as downtime, and sends SMS, email, Slack and webhook alerts on every plan with no credit packs. Flat pricing from $19 a month, with 30-second checks from $59, and no execution meter anywhere on the invoice.

See pricing

keep reading