ScaleOps Pricing: What ScaleOps Costs on AWS Marketplace and How It Compares to StormForge and CAST AI

ScaleOps publishes no price on its own website, but it does publish one on AWS Marketplace. As of 16 September 2026, scaleops.com/pricing is a quote form that says the pricing model depends on factors specific to your environment. The ScaleOps Platform listing on AWS Marketplace (listing ID prodview-t6ctoxyyxp5zs, sold by ScaleOps) publishes a fixed platform fee of $4,167 for one month or $50,000 for twelve months, plus a metered fee of $9.00 per vCPU, and states that ScaleOps does not currently support refunds. That is the only first-party ScaleOps price in existence, and almost nobody comparing rightsizing vendors has read it.

That matters because the two most-cited sources on ScaleOps pricing say the opposite. CloudBolt, which owns StormForge, runs a ScaleOps vs StormForge comparison table with the entry "Custom quote only. No public pricing. 7-day trial." Software directories repeat that ScaleOps has a single custom-quoted tier. Both are correct about the vendor site and wrong about the vendor. The Marketplace listing is a public offer that any AWS account can subscribe to, and its numbers are the anchor you negotiate from.

Everything below was read off ScaleOps's site, CloudBolt's published comparison and the three AWS Marketplace listings involved (ScaleOps, StormForge and CAST AI) on 16 September 2026. Where a number comes from a competitor describing ScaleOps rather than from ScaleOps, this page says so.

How much does ScaleOps cost?

On the public AWS Marketplace rate card, ScaleOps costs $50,000 a year in platform fees plus $9.00 per vCPU it manages. The listing structures this as a contract with usage. The contract part is a single dimension called ScaleOps Platform Fee, described as "A fixed rate for ScaleOps platform", priced at $4,167 for a one-month term and $50,000 for a twelve-month term. The usage part is a dimension called ScaleOps Platform Metered Fee, described as "Charge Per vCPU", priced at $9.00 per unit. The listing carries an auto-renew badge and the support term reads, verbatim, "we do not currently support refunds, but you can cancel at any time."

Two things stand out once you read the card rather than the summary. First, there is no annual discount. Twelve monthly terms at $4,167 come to $50,004, so the yearly contract saves you four dollars. Second, the platform fee is charged before a single vCPU is metered. A 200-vCPU cluster and a 2,000-vCPU estate both start at $50,000 a year; the metered fee is what separates them. That shape favors large estates and punishes small ones, which is why CloudBolt can price StormForge pay-as-you-go with no minimum and win the small-cluster deals without ever competing on the per-vCPU rate.

What the listing does not say is the metering period for the $9.00. AWS Marketplace usage dimensions are reported by the vendor, and the description "Charge Per vCPU" gives no interval. If it is per vCPU per month, a 1,000-vCPU estate pays $108,000 a year in metered fees on top of the platform fee, for $158,000 in total. If it is per vCPU per contract term, the same estate pays $59,000. Ask ScaleOps which it is before you put a number in a budget request, and get the answer in the order form, not on a call.

Is ScaleOps free?

No. ScaleOps offers a free trial, and then it charges the platform fee plus the per-vCPU meter. The trial length is a competitor's claim rather than the vendor's: CloudBolt's comparison says seven days, and ScaleOps's own site says only "Get Started" and "Quick Start" without stating a duration. There is no free tier and no free plan on the Marketplace listing. If a genuinely free entry point matters, PerfectScale publishes a free tier up to 200 vCPU a month with unlimited clusters, CAST AI's Marketplace listing includes a $0 monitoring plan, and the Kubernetes Vertical Pod Autoscaler is open source. None of those is ScaleOps, and none of them manages nodes, spot capacity and GPUs the way ScaleOps does.

What is on the ScaleOps AWS Marketplace rate card?

That is the whole card. There is no tier ladder, no per-cluster cap and no bundled support level, which makes it simpler than CAST AI's card and less flexible than StormForge's. Quantity configuration is disallowed on the contract dimension, so you cannot buy two platform fees to cover two business units; you buy one and the meter does the rest.

Why does scaleops.com say there is no public price?

Because the website is written for the sales motion and the Marketplace listing is written for procurement. The quote form on scaleops.com uses almost the same sentence CAST AI uses on its pricing page, that the pricing model depends on factors specific to your environment. For an enterprise deal that will be closed with a private offer, that sentence is true: the public card is a ceiling, and the private offer is where the discount lives. For anyone trying to build a budget before a sales call, it is unhelpful, and the Marketplace listing is where the real starting point sits. The same pattern holds for Finout, whose Business tier is $12,000 a year on Marketplace and "request a quote" on its own site, and for CloudHealth and Cloudability, whose Marketplace ladders are the only public prices either has ever carried.

How does ScaleOps pricing compare to StormForge and CAST AI?

All three publish an AWS Marketplace rate card, and the shapes are different enough that the cheapest vendor depends entirely on your size. The full head-to-head on architecture, in-place resizing and scope lives in the ScaleOps vs StormForge comparison; this table is the price layer only.

Run the arithmetic at three sizes. At 300 vCPUs, StormForge pay-as-you-go costs about $10,800 a year, CAST AI Growth costs $12,000 plus $18,000 in CPU-hours for about $30,000, and ScaleOps costs $50,000 before its meter is applied. At 1,000 vCPUs, StormForge on contract costs $32,400, CAST AI GrowthPro costs about $72,000, and ScaleOps costs $50,000 plus metering, which is at least $59,000 and, if the meter is monthly, $158,000. At 5,000 vCPUs, StormForge costs $162,000, CAST AI Enterprise costs $60,000 plus $300,000 in CPU-hours for $360,000, and ScaleOps costs $50,000 plus $45,000 or $540,000 depending on that one unanswered question about the metering period. The numbers say what the listings say: StormForge is the cheapest option up to a few thousand vCPUs and has the lowest per-vCPU rate on any card, CAST AI has the smallest platform fee but a meter nearly double StormForge's, and ScaleOps is a large fixed fee whose total cost you cannot know until the vendor tells you what "per vCPU" means. At 5,000 vCPUs that one answer decides whether ScaleOps is the cheapest vendor on the page or the most expensive.

Price per vCPU is not the whole decision, because the three products do different amounts of work. StormForge rightsizes pod requests and stops there; CloudBolt states GPU support is on the roadmap and spot optimization is not offered. CAST AI and ScaleOps manage the node layer as well, and ScaleOps adds GPU rightsizing and an air-gapped deployment that neither of the others offers. If you only want pod rightsizing, you are paying ScaleOps for capabilities you will not switch on. If nobody on your team owns node management, the platform fee buys you the vendor that does.

What does the ScaleOps platform fee actually buy?

The scope on ScaleOps's site is unusually wide for the category: continuous pod rightsizing, HPA optimization, minimum and maximum replica tuning, node optimization and consolidation, Karpenter optimization, spot instance handling, smart pod placement, fractional and autonomous GPU workload rightsizing, and Java resource management, all fully autonomous from install. Its published results are Maxar, now Vantor, at 62 percent CPU savings and 40 percent memory savings, and Outbrain at over 40 percent cost savings, with a headline claim of cutting costs by up to 80 percent. It runs self-hosted, in ScaleOps Cloud, or air-gapped, and CloudBolt's comparison states that it requires a full in-cluster Prometheus deployment per cluster, which is a real infrastructure line to add to the $50,000. Node consolidation is also the part of the platform with a side effect finance rarely models: fewer nodes running around the clock lowers the estate's electricity draw, which is why some FinOps teams now hand the before-and-after node counts to whoever keeps the company's carbon accounting current.

What the fee does not buy is an independent measurement of the savings. ScaleOps is paid on the vCPUs it manages and reports its own results, which is not a criticism, it is the business model of every automation vendor on this page. The practical fix is to hold a read-only view of Kubernetes cost per team and per product that ScaleOps does not touch, take a baseline during the trial, and check the savings claim against the bill finance actually pays. That is the one job Costanalyst does in this evaluation: read-only, no controller in the cluster, published pricing, Kubernetes cost data on the Scale plan.

What should I ask ScaleOps before signing?

Six questions, all of them answerable in writing before a private offer is drafted.

1. What period does the $9.00 per vCPU cover? Per month, per term, or per hour-equivalent. This single answer moves a 1,000-vCPU quote by roughly $100,000 a year.
2. Which vCPUs are metered? Every vCPU on a node the agent can see, or only the vCPUs of workloads under active optimization. Ask whether system namespaces, DaemonSets and excluded workloads count.
3. Is the private offer discounting the platform fee, the meter, or both? The public card has no annual discount, so any discount you are offered is against list; make sure you know which list.
4. What happens at the auto-renewal? The listing auto-renews and does not support refunds. Get the renewal notice period and the price-increase cap into the order form.
5. Who runs the Prometheus? If the deployment is self-hosted with a Prometheus per cluster, ask for a sizing guide for your workload count and add the storage and compute to your total cost.
6. What is the trial length and scope? The vendor site does not state it; CloudBolt says seven days. Ask for enough time to cover one full weekly cycle on a namespace you understand, and run a read-only cost baseline alongside it.

The short version

ScaleOps's website publishes no price; its AWS Marketplace listing does: $50,000 a year or $4,167 a month in platform fees, plus $9.00 per vCPU with the metering period unstated, no refunds, auto-renew. StormForge has the lowest per-vCPU rate and no minimum on pay-as-you-go, CAST AI has the smallest platform fee and the most expensive meter, and ScaleOps is the only one of the three that manages GPUs and runs air-gapped. If you are shortlisting, the ScaleOps alternatives guide covers the full field, the Kubernetes rightsizing tools comparison sorts every vendor by whether it restarts pods to resize them, and the CAST AI pricing and pod rightsizing pricing pages carry the rest of the rate cards. Take a baseline before the trial, and make the savings number come from something the vendor does not invoice.