Cheapest cloud compute per vCPU-hour

Scaling out horizontally, where unit compute price is what compounds.

List prices checked

How this is scored

Ranked purely on published vCPU-hour rate. Flat-rate hosts that do not quote an hourly price are excluded rather than converted, since the conversion would be our arithmetic rather than their price.

All 10, ranked for this workload

#ServiceProviderCost for this workloadNotes
1Vultr Cloud ComputeVultr$0.004/vCPU-hr · $2.92/mo continuousNone; the $2.50 IPv6-only plan is the entry point
2Hetzner CloudHetzner$0.0052/vCPU-hr · $3.80/mo continuous20 TB of traffic included with every server
3DigitalOcean DropletsDigitalOcean$0.00595/vCPU-hr · $4.34/mo continuous500 GB to 1 TB of pooled egress per Droplet
4Scaleway InstancesScaleway$0.006/vCPU-hr · $4.38/mo continuous—
5UpCloudUpCloud$0.0074/vCPU-hr · $5.40/mo continuous—
6Linode (Akamai Cloud)Akamai$0.0075/vCPU-hr · $5.47/mo continuous1 TB of egress included on the $5 plan
7Oracle Cloud ComputeOracle$0.01/vCPU-hr · $7.30/mo continuous4 Arm cores, 24 GB RAM, and 10 TB egress per month, always free
8AWS EC2Amazon Web Services$0.0416/vCPU-hr · $30/mo continuous750 hours per month of a t-class micro instance for 12 months
9Google Compute EngineGoogle$0.0475/vCPU-hr · $35/mo continuousOne e2-micro instance per month, always free
10Azure Virtual MachinesMicrosoft$0.052/vCPU-hr · $38/mo continuous750 hours per month of a B1s instance for 12 months

Published list prices only. Rows that do not publish a rate for an axis this workload depends on are excluded rather than shown as cheap.

Questions

Cheapest cloud compute per vCPU-hour in 2026?
Vultr Cloud Compute from Vultr — $0.004/vCPU-hr · $2.92/mo continuous. Hetzner Cloud is second at $0.0052/vCPU-hr · $3.80/mo continuous. This ranking is specific to the workload described above; a different usage shape reorders it.
How is this ranking weighted?
Ranked purely on published vCPU-hour rate. Flat-rate hosts that do not quote an hourly price are excluded rather than converted, since the conversion would be our arithmetic rather than their price.
Why does this differ from the general cheapest list?
Because a single price axis never describes a real workload. Ranking by headline rate answers "who is cheapest per unit"; this page answers "who is cheapest for this job", and the two orders are often very different — which is the whole reason it exists as a separate table.
What is not accounted for?
Committed-use discounts, regional price variation, per-request charges, support plans, and minimum retention. These are published list prices applied to one stated workload — a shortlist to price properly with the vendor, not a quotation.

Other workloads