Skip to main content

CubePath VPS benchmarks

Benchmarking right now

Tested plans

3 plans tested out of 18 active. Scores are medians over the last 90 days.

Plan vCPU RAM Price Runs GB7 single GB7 multi Last tested
gp.micro 2 shared vCPU 4 GB $8.11/mo
2 runs
1,195 2,176 Oct 13, 2026
rz.starter 2 shared vCPU 4 GB $16.22/mo
1 run
2,563 4,597 Oct 12, 2026
gp.starter 4 shared vCPU 8 GB $15.21/mo
1 run
1,207 4,134 Oct 12, 2026

Geekbench 7 single-core

Average per plan, higher is better

Loading...

Disk 4K mixed IOPS

Average per plan, higher is better

Loading...

Performance by location

Every CubePath location

Medians over the last 90 days. Each cell is the Geekbench 7 single-core median, with 4K IOPS under it. Click a location name for every plan and result there.

Plan Barcelona Houston Miami Price Multi per $
gp.micro n/a 1,235
7,809 IOPS
1,154
7,853 IOPS
$8.11/mo
268
$3.73 per 1,000 points
rz.starter 2,563
10,262 IOPS
n/a n/a $16.22/mo
283
$3.53 per 1,000 points
gp.starter 1,207
7,818 IOPS
n/a n/a $15.21/mo
272
$3.68 per 1,000 points

Click a score to open it in the side-by-side. Multi per $ is the multi-core median divided by the current monthly price, USD only. Each plan's figure is the average across its tested locations. Bars are relative to CubePath's best plan.

A location is highlighted only when it leads the next best by 5 percent or more. Smaller gaps are normal run-to-run variance.

Consistency across locations

Whether a CubePath plan performs the same no matter which location you pick.

Mostly consistent across locations 1 plan tested in two or more locations: 1 consistent.

We don't have enough data to be all that confident in these yet. Most locations have only one or two runs so far, and two runs in the same location can land 20 percent apart on Geekbench and more than 50 percent apart on disk. With that few runs, some of the gaps below are just run-to-run noise, not the location. They'll firm up as each location gets 3 or more runs.

Plan Rating GB7 single gap GB7 multi gap 4K IOPS gap CPU
2 runs in 2 locations
Consistent
One run per location
7%
Slowest: Miami
9%
Slowest: Miami
1%
Slowest: Houston
2 different models
AMD EPYC Processor, Intel Core Processor (Broadwell, no TSX, IBRS)

Each figure is how far the slowest location's median sits below the fastest location's median, over the last 90 days. Geekbench is rated consistent under 10 percent and inconsistent over 25 percent. Disk is noisier from run to run, so 4K IOPS gets 25 percent and 50 percent. A plan takes its worst rating.

Tested in one location only, so not rated yet: gp.starter and rz.starter.

4 runs, newest first.

Benchmark runs
Server Specs Storage Network Single-core Multi-core 4K IOPS Monthly price Tested
CubePath
gp.micro
Houston
1,235
2 vCPU 4 GB 80 GB NVMe
$8.11/mo
CubePath
rz.starter
Barcelona
2,563
2 vCPU 4 GB 80 GB NVMe
$16.22/mo
CubePath
gp.micro
Miami
1,154
2 vCPU 4 GB 80 GB NVMe
$8.11/mo
CubePath
gp.starter
Barcelona
1,207
4 vCPU 8 GB 100 GB NVMe
$15.21/mo

18 active plans in 4 product lines.

Provider details

Website
cubepath.com/cloud-vps
Port 25
On request
IPv4
Included
IPv6
Included
Plan data
Synced from API
Locations
3

About CubePath

We create CubePath servers through their API, run YABS with the same flags as every other provider, and delete the server as soon as results are saved. Plan and price data is synced from the CubePath API, last synced Oct 14, 2026.