Cloudways vs Kinsta: Server Hours or Visitor Meters?

Cloudways vs Kinsta is not really a contest between two feature lists. It is a contest between two meters. Cloudways bills the server you rent, by the hour, and does not care how many people visit. Kinsta bills a plan per account, metered by monthly visits or by server bandwidth, and does not care how much RAM your sites use. Which one costs less depends almost entirely on the shape of your traffic and how many sites you run.

So instead of another feature checklist, I priced five realistic setups on both hosts, using the live pricing pages and billing docs I read on 17 October 2026. All prices below are US dollars per month, billed monthly, before tax. The short version of this Cloudways vs Kinsta comparison: one small site is close to a tie. A single viral month costs less on Kinsta than its reputation suggests. A portfolio of small sites is where Cloudways wins by a wide margin.

What each meter actually counts

On Cloudways Flexible, you pick a cloud provider and a server size, and the pricing page lists every size with “Unlimited Visits” and “Unlimited Websites”. A 2 GB DigitalOcean Premium server is $28 a month, shown as $0.0417 an hour. The 4 GB size is $54 ($0.0804 an hour) and the 8 GB size is $99. Cloudways charges in arrears: its billing FAQ says invoices go out in the first week of each month for the previous month’s usage. The only traffic-linked charge is bandwidth beyond the server’s allowance (2 TB on the 2 GB server), billed at $0.02 per GB on DigitalOcean.

On Kinsta, the meter is shared across every site in your plan, and you choose which meter it is. Its visit-counting doc defines a visit as each unique IP address seen in a 24-hour period. Well-known bots are filtered out, visits to staging are ignored, but visits served from the CDN or edge cache still count. The alternative is a bandwidth plan, where Kinsta’s plan docs say only data sent from the origin server counts and “anything served by a CDN or edge cache is excluded.”

Go over either allowance and the overage page charges $0.50 per 1,000 visits or $0.50 per GB.

That second meter matters, because most of the Kinsta vs Cloudways coverage I read, including Cloudways’ own Kinsta alternative page, leads with the visit overage. Kinsta’s bandwidth plans cost the same as the visit plans at each tier, so a site can opt out of counting visitors entirely.

What moves the bill	Cloudways Flexible	Kinsta
More visitors	Nothing, until the server needs more CPU or RAM	Visit plans: $0.50 per 1,000 over the cap
More data from the origin	$0.02/GB past the transfer allowance (DigitalOcean)	Bandwidth plans: $0.50/GB over the cap
Another WordPress site	Nothing, if the server has room	$30 a month per extra site, or a bigger plan
A bigger server	The next size’s hourly rate	Not billed: on standard plans each live site container gets 12 CPUs and 8 GB of RAM by default
When you pay	Hourly usage, invoiced monthly in arrears	Monthly or annual in advance; annual gives two months free

That RAM line comes from Kinsta’s technical FAQ. It explains the whole Cloudways vs Kinsta pricing split. Kinsta sells capacity per site and absorbs the hardware question, while Cloudways sells hardware and lets you pack as many sites onto it as it will hold.

Five Cloudways vs Kinsta bills, priced side by side

1. A brochure site with 20,000 visits a month

Kinsta’s Single 35k plan covers it at $35, or $350 a year if you pay annually, which works out to about $29.17 a month. On Cloudways, the 1 GB server ($14 Premium, $11 Standard) looks tempting. But Cloudways’ Getting Started FAQ says 1 GB servers “are primarily intended for testing and staging purposes” and advises 2 GB or more for live sites. That makes the honest comparison $28 against $35. Cloudways’ backups, CDN and email are extras, and once you add them the gap mostly closes. I price those add-ons line by line in my Cloudways review, so I won’t repeat them here. At this size, the Cloudways vs Kinsta choice should rest on support and convenience, not price.

2. The same site in a viral month: 100,000 visits

This is the scenario Cloudways’ marketing is built on, and the numbers are milder than the pitch suggests. Staying on Single 35k, the extra 65,000 visits cost $32.50, so the month comes to $67.50. Kinsta’s own docs use the same example. Kinsta’s $35 Launch plan meters 15 GB of server bandwidth instead of visits (Kinsta estimates about 75,000), so a mostly cached site should price that too. On Cloudways, the bill stays at $28 if the 2 GB server copes. If it doesn’t, Cloudways’ scaling guide lets a DigitalOcean server scale CPU and RAM only, which it calls “ideal for short-term traffic surges”. You can revert later, though only back to the size you launched with. A week at the 4 GB hourly rate instead of the 2 GB rate is about $6.50 extra by my arithmetic from the listed rates. Cloudways shows the exact price when you scale.

So in a spike month, Cloudways comes out $33 to $39.50 cheaper. The catch is who does the work. On Kinsta the spike is a line on an invoice. On Cloudways it is a server you have to watch and resize in time, or the site slows down.

3. A WooCommerce store that grows to 120,000 visits

Kinsta marks Single 65k and every larger plan as recommended for ecommerce. At 120,000 steady visits, Single 125k at $90 is the right fit. Cloudways’ cards describe the 4 GB server ($54) as for “more demanding websites with higher traffic” and the 8 GB server ($99) as “ideal for e-com stores”. So a realistic Cloudways range is $54 to $99 depending on how much uncached cart and checkout traffic the store has. Kinsta’s Launch Plus ($59, 25 GB of server bandwidth, an estimated 125,000 visits) is also marked for ecommerce. So if the store’s origin traffic fits 25 GB, the Cloudways vs Kinsta price difference runs from $36 in Cloudways’ favour to $40 in Kinsta’s. One warning for anyone converting numbers: Cloudways sizes servers in page views, while Kinsta counts unique IPs per day, and one visit can be many page views. If you run a store, my WooCommerce development work starts with exactly this sizing question.

4. Ten small client sites, 80,000 visits combined

This is where the two meters diverge most. Kinsta’s cheapest ten-site plan is Scale at $169, which meters 125 GB of server bandwidth (Kinsta estimates about 650,000 visits). On the visit meter it is WP 10 at $225, or WP 5 ($115) plus five site add-ons at $30 each, which is $265. On Cloudways, its own FAQ says a 2 GB server “can handle 4-5 small websites having combined web traffic up to 110K page views per month”. So ten small sites fit on two 2 GB servers ($56) or one 4 GB server ($54). Per-domain extras do grow with the site count (Cloudflare Enterprise starts at $4.99 per domain a month). Even with ten of those, Cloudways comes to about $104 to $106 a month, against $169 to $225 on Kinsta. For agencies, this scenario usually settles the Cloudways vs Kinsta question by itself.

5. A heavily cached publication at 600,000 visits

On a visit plan, Kinsta’s Single 750k costs $375. This is the Cloudways vs Kinsta case where Kinsta’s second meter decides the answer.

At every tier the two plan types cost the same, so the question becomes how much data each visit pulls from the origin. Single 125k ($90) and Single 65 GB ($90) break even at about 0.52 MB of origin traffic per visit.

For 600,000 visits, Single 125 GB at $170 holds as long as the origin sends less than about 208 KB per visit. That is plausible when the CDN and edge cache serve the images, scripts and most of the HTML. You can check your current origin bandwidth in your host’s logs before choosing. On Cloudways this site is purely a server-sizing question. Only a load test against your real cache hit rate will tell you whether 4 GB or 8 GB is enough.

Two rules the sales pages skip

Kinsta’s overage has a ceiling on patience. Paying $0.50 per 1,000 extra visits is cheap enough that staying on a small plan can look smarter than upgrading. Kinsta’s overage page closes that loophole with an “extreme overage” rule. When the overage equals or exceeds the plan price, or $500, whichever is less, Kinsta may demand immediate payment or an upgrade, and “reserve[s] the right to temporarily limit access to the site”. On Single 35k that line sits at 105,000 visits. Upgrading mid-month doesn’t wipe the overage either; Kinsta says new plans are “not applied retroactively”.

Cloudways scaling is not always reversible. The CPU-and-RAM-only resize that makes the spike scenario cheap is a DigitalOcean feature. Scaling CPU, RAM and storage together “is permanent”, and on Vultr, Cloudways says scaling is permanent, so going back down means cloning to a smaller server. AWS and Google Cloud servers can scale back down. If you expect seasonal peaks, choose the provider before you launch, not after. Both rules change the Cloudways vs Kinsta maths for spiky sites more than any feature table does.

What both status pages showed me this week

I can’t run a fair Cloudways vs Kinsta speed test without client sites on both hosts, so for a first-hand check I read both companies’ public incident histories. On 17 October 2026 at 12:22 UTC I pulled the incident feeds from status.cloudways.com and status.kinsta.com (both run on Atlassian Statuspage, which exposes them as JSON). I counted everything opened since 19 July, roughly the previous 90 days.

• Cloudways: 13 incidents. Cloudways marked two as critical. One was a Google Cloud server disruption across multiple regions on 11 August (about five and a half hours from first post to resolved), the other a Rackspace webmail problem. Most of the rest involved add-ons (Rackspace and Elastic Email, DNS Made Easy) or upstream provider APIs. That includes a Vultr API degradation that stayed open for 13 days and affected launches, clones and reboots rather than running sites. One more, a two-hour DigitalOcean connectivity outage in Bangalore on 19 July, did affect running sites in that data center.
• Kinsta: 10 incidents. Six were service disruptions at a single data center: Northern Virginia twice, Toronto, Phoenix, Frankfurt and Eemshaven, mostly resolved within three to eight hours. The rest were dashboard and support-system issues, plus the plugin notice below.

Two caveats keep this honest. The counts reflect how each company reports as much as how often things break. Kinsta lists 58 status components, mostly individual data centers, while Cloudways lists 17, mostly whole cloud providers and add-ons. And the durations run from first post to “resolved”, which is not the same as downtime for any given site. One detail I liked: on 20 August both hosts posted the same warning about a WP Rocket update causing fatal errors on WordPress 7.1, within hours of each other. If that kind of error ever takes your site down, my guide to the WordPress white screen of death walks through the fix.

On the “is Kinsta safe” question people ask, the pricing page lists Cloudflare’s WAF and DDoS protection, bot protection and free malware removal on every plan. Cloudways includes an Imunify360 server firewall and sells malware protection from $4 per app a month. Neither replaces keeping plugins updated, which my WordPress security guide covers.

How I would choose

The links to both hosts below are affiliate links. If you sign up through them I may earn a commission, at no extra cost to you. The affiliate disclosure page explains how I handle that.

My Cloudways vs Kinsta rule of thumb comes down to two questions. Run three or more sites, or have someone who will read CPU graphs and resize a server? Cloudways charges for hardware, and hardware is cheap to share. Its 3-day trial needs no credit card, which is enough to clone a site and load-test it.

Run one or two sites and want the spike, the server and the malware cleanup to be someone else’s job? Kinsta costs more for small portfolios, but the price is the whole job. Single 35k and Launch include the first month free, and a 30-day money-back guarantee applies after any free period. If your site is mostly cached, compare the bandwidth plan before the visit plan.

If you are weighing Kinsta against its other premium rival instead, my Kinsta vs WP Engine comparison covers that matchup. If the budget is closer to shared hosting, start with Hostinger vs SiteGround. Whichever way you move, my WordPress migration service handles the switch and the post-move checks. Speed optimization is usually cheaper than buying a bigger plan to hide a slow site.