Splunk Observability Cloud is priced per host, in three bundles billed annually: Infrastructure from $15 per host a month, App & Infra from $60 and End-to-End from $75. APM is no longer listed as a standalone price. It comes inside App & Infra and End-to-End. Synthetic uptime monitoring is included in all three bundles, API synthetics need App & Infra, and browser synthetics and real user monitoring need End-to-End. A 20-host estate on App & Infra runs $1,200 a month, or $14,400 a year.
The flat per-host structure is the important part, and it is unusual in this category. A 2 GiB container node and a 512 GiB database server cost exactly the same $60. That makes Splunk predictable in a way its main rival is not, and it is the single biggest reason a large-server estate ends up cheaper here than on a memory-metered platform.
Splunk sells two APM products, and the prices are four times apart
Before any of the numbers below are useful, you need to know which product you are being quoted. Cisco acquired AppDynamics, then acquired Splunk, and appdynamics.com/pricing now 301-redirects to Splunk's observability pricing page. So Splunk has two application monitoring products on one rate card, billed on completely different axes.
Splunk Observability Cloud APM comes in the App & Infra bundle at $60 per host a month. Splunk AppDynamics is priced per virtual CPU, from $6 a month on Infrastructure Edition to $33 on Premium and $50 on Enterprise. On an 8 vCPU machine, AppDynamics Premium is $264 a month and Observability Cloud App & Infra is $60 for the same box. That is not a negotiable discount. It is a different SKU, and which one an account team leads with tends to depend on which part of the portfolio the conversation started in.
Ask for both in writing. If your servers are large and few, the per-host SKU is dramatically cheaper. If you run many tiny single-core instances, the per-vCPU SKU can win. The full per-core breakdown is in our AppDynamics pricing breakdown.
The Splunk Observability Cloud rate card
These are the US dollar list prices published on Splunk's own pricing page. Every figure is a "starts at" number and everything is billed annually, so treat list as a ceiling and a way to compare shapes rather than as a quote.
| Capability | List price | Billed as |
|---|---|---|
| Infrastructure bundle (infrastructure monitoring, Log Observer Connect, Network Explorer, synthetic uptime monitoring) | From $15/mo | Per host, billed annually |
| App & Infra bundle (adds APM with Always On Profiling and synthetic API monitoring) | From $60/mo | Per host, billed annually |
| End-to-End bundle (adds real user monitoring and synthetic browser monitoring) | From $75/mo | Per host, billed annually |
| Database Monitoring, Secure Application | Contact sales | Optional add-ons |
| AppDynamics browser synthetics | From $12/mo | Per test location, billed annually |
| Splunk AppDynamics | $6 to $50/mo | Per vCPU, billed annually |
Two things are worth noticing. First, the only per-location synthetic price on the page is labeled as an AppDynamics add-on. On the Observability Cloud side, synthetics are no longer metered on the published card at all: they are features of a bundle, and the bundle decides which check types you get. Second, there is no published seat price anywhere. Adding engineers to the console does not move the bill, which is a real advantage over New Relic, where full platform users are billed individually.
What counts as a host, and why it decides your bill
A host is a monitored server or instance, and the price does not vary with its size. That single design choice is what makes Splunk cheap on big machines and relatively expensive on small ones.
Compare it against Dynatrace, which bills Full-Stack Monitoring at $0.01 per memory-GiB-hour, or $7.30 per gigabyte a month. Divide Splunk's $60 App & Infra bundle by $7.30 and the two meet at 8.22 GiB of RAM per host. Below that, Dynatrace is cheaper. Above it, Splunk is, and because one price is flat and the other is a straight line, the gap widens with every gigabyte. A 64 GiB database server is $60 on Splunk and $467.20 on Dynatrace. We worked the whole curve through in Dynatrace vs Splunk pricing. If your estate is mostly network gear rather than servers, the per-device comparison in LogicMonitor vs Splunk pricing is the more useful one, because Splunk's host card does not price switches or access points. And if the other name on your shortlist is Datadog, our Datadog vs Splunk pricing comparison shows why APM favors Datadog while autoscaling fleets favor Splunk.
The practical consequence: if you run container fleets with 2 GiB nodes, per-host billing is the expensive model and you should price the alternatives. If you run a few dozen substantial application and database servers, per-host billing is the cheapest structure in this category.
Synthetic monitoring comes inside a bundle, not on a meter
Splunk's pricing page no longer publishes a per-request synthetic rate. Synthetic Uptime Monitoring is listed as included in all three bundles, and the page gives no per-check allowance and no overage rate. So the cost of an uptime check is the cost of the host bundle it rides on, from $15 per host a month billed annually. API checks need App & Infra at $60 per host, and browser checks need End-to-End at $75 per host. Dynatrace, by contrast, still meters HTTP monitors at $1.00 per 1,000 requests, so a one-minute check from three locations is $131.50 a month there.
| What you run | Splunk bundle needed | Published starting price |
|---|---|---|
| HTTP uptime checks | Infrastructure (any bundle) | $15 per host a month |
| API checks | App & Infra or End-to-End | $60 per host a month |
| Browser journeys | End-to-End | $75 per host a month |
| Real user monitoring | End-to-End | $75 per host a month |
| Browser journeys on AppDynamics | AppDynamics add-on | $12 per test location a month |
Browser tests on the AppDynamics side work differently and the difference matters. At $12 per test location a month, flat, you can run a checkout journey every minute or once an hour for the same money. Dynatrace meters the same job at $0.0045 per synthetic action, which makes a single-action script at one-minute frequency $197.25 and a five-step checkout $986.24. Flat billing wins comfortably at any frequency a real team would use.
The catch is the floor underneath all of it. Synthetic monitoring is part of an Observability Cloud bundle, not a standalone product, so the real entry price is the host fee: $15 a host a month at minimum, billed annually, before the first check runs. Because no per-check allowance is published, ask in writing how many checks and locations your bundle covers. And the fastest interval is one minute. Splunk AppDynamics documentation gives one minute as the minimum frequency for API monitoring, and Dynatrace, Datadog and New Relic all share that floor.
Three worked bills
List prices, billed annually, App & Infra bundle (the cheapest one that includes APM), using each vendor's published rate card. The Dynatrace column prices Full-Stack Monitoring at $0.01 per memory-GiB-hour over a 730-hour month, with no synthetics or security add-ons on either side, so the two columns compare like for like.
| Estate | Splunk App & Infra | Dynatrace equivalent |
|---|---|---|
| Startup: 5 hosts, 4 GiB each | $300.00/mo | $146.00/mo |
| Mid-size: 20 hosts, 32 GiB each | $1,200.00/mo | $4,672.00/mo |
| Database-heavy: 8 hosts, 64 GiB each | $480.00/mo | $3,737.60/mo |
The middle row is the one to sit with. Twenty ordinary 32 GiB application servers is $1,200 a month on Splunk and $4,672 on Dynatrace, a difference of $41,664 a year on identical hardware, driven entirely by the billing unit. The top row shows the reverse: on small 4 GiB machines Dynatrace is roughly half the price. Neither vendor is simply cheaper, and any write-up that tells you one is has not done the arithmetic.
What actually drives the bill up
Host count is the obvious lever and the one most teams already watch. The three that surprise people are quieter. Ephemeral infrastructure is the first: autoscaling groups and short-lived build agents each register as hosts, and a fleet that peaks at 60 instances does not bill like the 20 you picture. The second is bundle creep: adding browser checks or real user monitoring means moving from App & Infra at $60 to End-to-End at $75, and because bundles are priced per host, that upgrade multiplies across the whole estate rather than adding a flat fee. The third is the second APM SKU discussed above quietly appearing on the same quote.
Because the meter is per host and the invoice arrives annually, the gap between what you were quoted and what you actually run tends to go unnoticed for a full renewal cycle. The fix is unglamorous: reconcile the vendor's host count against your own inventory every month, the same way you would track cloud and SaaS spend read-only against what you actually provisioned, so the true number is a fact you already know rather than something a renewal conversation reveals.
What the host fee does not buy you
Splunk Observability Cloud is a serious platform and this is not an argument against it. But two gaps come up repeatedly with teams who bought it expecting uptime monitoring to be handled.
The first is detection speed. One minute is the fastest synthetic interval at any price, and with a single-location check you are also trusting one vantage point, so a regional network problem reads as an outage. The second is who finds out. Escalation to a named person by SMS, a public status page your support team can point customers at, and confirmation from a second region before anyone gets woken at 3am are not observability platform features. Our API monitoring with status code, keyword and JSON path assertions covers that job on flat plans from $49 a month, at 30-second resolution from $59, and the honest case for running both is laid out in uptime monitoring to pair with APM, including the situations where a second tool is plainly not worth it.
Is Splunk Observability Cloud worth the price?
For a mid-size estate of substantial servers, the per-host model is the most predictable pricing in the category and it beats memory-metered billing badly on large machines. For container fleets of small nodes it is the expensive structure and worth pricing against Dynatrace. For teams that only want to know whether the site is up, it is the wrong shape entirely: you are buying an agent-based platform with a host floor and a one-minute interval to answer a question a flat external monitor answers in 30 to 60 seconds. Compare the whole field in our best uptime monitoring tools roundup and the uptime monitoring pricing comparison.
Questions people ask about Splunk Observability Cloud pricing
How much does Splunk Observability Cloud cost per host?
Splunk sells three bundles, all per host per month and billed annually: Infrastructure from $15, App & Infra from $60 (the cheapest bundle with APM) and End-to-End from $75. The price does not change with the size of the host, so a 2 GiB node and a 512 GiB database server cost the same.
Does Splunk Observability Cloud have a free tier?
No permanently free tier is published. Splunk offers trials of its observability products, but there is nothing comparable to the New Relic free tier, which includes one full user, 100 GB of monthly ingest and unmetered simple ping monitors. Every Splunk figure listed is a paid, annually billed rate.
Is Splunk cheaper than Dynatrace?
On large hosts, clearly yes. Splunk App & Infra, which includes APM, is a flat $60 per host and Dynatrace Full-Stack is $7.30 per GiB of RAM a month, so the two meet at 8.22 GiB. Above that ratio Splunk is cheaper and the gap grows: a 64 GiB server is $60 against $467.20. Below it, on container-sized nodes, Dynatrace is cheaper.
How much does Splunk synthetic monitoring cost?
Synthetic uptime monitoring is included in every Observability Cloud bundle, from $15 per host a month billed annually, and Splunk's pricing page publishes no per-check allowance or overage rate. API checks need App & Infra at $60 per host and browser checks need End-to-End at $75. On the AppDynamics side, browser synthetics are an add-on at a flat $12 per test location a month. Either way the host fee applies before any check runs.
What is the fastest check interval in Splunk synthetic monitoring?
One minute. Splunk AppDynamics documentation gives one minute as the minimum frequency support for API monitoring. Dynatrace, Datadog and New Relic share the same floor, so one minute is a category limit rather than a Splunk restriction, and no contract tier changes it.
Does Splunk Observability Cloud charge per user?
No seat price is published. Pricing is per host, so adding engineers to the console costs nothing. Dynatrace works the same way. New Relic is the outlier in this group, billing full platform users individually, which is what makes its sixth seat unusually expensive.