Table of Contents
A plan against a meter
DigitalOcean's product is a Droplet: a size with memory, vCPUs, SSD and a transfer allowance at a monthly price that is the same in every region, billed by the second up to that price. Google Cloud's product is Compute Engine: a machine type in a region, billed per second, with a persistent disk, an external IP address and outbound transfer each on its own meter, discounted automatically for sustained use and further for commitments. The first tells you the bill before you start; the second tells you the rates. DigitalOcean also sells a metered line now (v5 Droplets, configured by resource and priced hourly), and Google gives away one small machine permanently, so each has borrowed a little of the other's model. The comparison that follows is of the two main products.
Side by side
| DigitalOcean | Google Cloud Compute Engine | |
|---|---|---|
| Pricing model | Fixed plans with disk and transfer; configurable v5 Droplets priced hourly | Machine type per region, per second, plus disk, IP and egress; sustained-use and committed-use discounts |
| Smallest printed price | Basic 512 MiB, $4 a month | Not printed as text (calculator); e2-micro free within limits |
| 2 GB class | Basic 2 GiB, 1 vCPU, 50 GiB, 2,000 GiB transfer, $12 | e2 shared-core types, priced per region in the calculator |
| Free tier | None printed on the pricing page | One e2-micro instance-month in us-west1, us-central1 or us-east1; 30 GB standard disk; 1 GB North American egress; $300 credit for 90 days |
| Billing granularity | Per second, 60-second minimum | Per second after a one-minute minimum |
| US locations | Five metros with standard Droplets (NYC, SFO all families; ATL1, RIC1, MKC1 Premium CPU only); MEM1 v5 and GPU only | Nine regions: Iowa, South Carolina, Ashburn, Columbus, Dallas, Oregon, Los Angeles, Salt Lake City, Las Vegas |
| Backups | 20% weekly or 30% daily of Droplet cost; snapshots $0.06/GB-month | Persistent-disk snapshots, per GB-month, priced on the disk pricing page (script-rendered) |
| Transfer | Allowance per Droplet (500 GiB at $4, 2,000 GiB at $12) | 1 GB free egress from North America in the Free Tier; otherwise per GB by destination |
| Platform around the VM | Managed databases, Kubernetes, App Platform, Spaces, load balancers | Cloud SQL, GKE, Cloud Run, Cloud Storage, BigQuery and the rest of Google Cloud |
| Windows | Not mentioned on the Droplet pricing page | On Google pages not read for this version |
DigitalOcean: what the pricing page prints
From the Droplet pricing page on 2026-09-19, Basic Droplets with the Regular CPU option: $4 (512 MiB, 1 vCPU, 10 GiB, 500 GiB transfer), $6 (1 GiB, 25 GiB, 1,000 GiB), $12 (2 GiB, 50 GiB, 2,000 GiB), $18 (2 GiB, 2 vCPUs, 60 GiB, 3,000 GiB), $24 (4 GiB, 2 vCPUs, 80 GiB), $48 (8 GiB, 4 vCPUs, 160 GiB) and $96 (16 GiB, 8 vCPUs, 320 GiB); dedicated-CPU families from $42 (CPU-Optimized), $63 (General Purpose), $84 (Memory-Optimized) and $131 (Storage-Optimized). Billing: "Bundled Plan and v5 Droplets are both billed on a per-second basis, with a 60-second minimum". Backups: "Percentage-based Backups, set at 20 % (Weekly) or 30 % (Daily) of Droplet cost", or usage-based from "$ 0.01 /GiB per month"; snapshots "$0.06/GB per month". Premium CPU options (which the docs say "use NVMe SSDs and have higher network throughput speed") and v5 prices are on other tabs or in the configurator and are not quoted. The DigitalOcean review has every family's ladder.
Google Cloud: what the documentation states
Google's Free Tier documentation lists, for Compute Engine, one non-preemptible e2-micro VM instance per month in us-west1 (Oregon), us-central1 (Iowa) or us-east1 (South Carolina), 30 GB-months of standard persistent disk, and 1 GB of outbound transfer from North America per month, with the limit counted "by time, not by instance"; plus a $300 credit that ends at $300 spent or 90 days, whichever first. Its Compute Engine pricing page states "After 1 minute, your instance is billed on a per-second basis" and an "automatic discount of up to 30% on resources that are used for more than 25% of a month", with committed-use discounts beyond that; the per-hour machine prices on that page are rendered by script and this page does not reproduce them. A paid small VM therefore has four lines (machine, disk, external IP, egress) whose total is a calculator answer, not a page answer. The Google Cloud review covers the Free Tier in full and the Google Cloud vs Lightsail page sets it against AWS's bundle.
US regions
Google Cloud: nine US regions per its regions and zones documentation, us-central1 (Council Bluffs, Iowa), us-east1 (Moncks Corner, South Carolina), us-east4 (Ashburn, Virginia), us-east5 (Columbus, Ohio), us-south1 (Dallas, Texas), us-west1 (The Dalles, Oregon), us-west2 (Los Angeles), us-west3 (Salt Lake City) and us-west4 (Las Vegas), with the free e2-micro limited to the first, second and sixth. DigitalOcean: per its regional availability page, standard Droplets are sold in five US metros, New York City (NYC1, NYC2, NYC3) and San Francisco (SFO2, SFO3) with every family, and Atlanta (ATL1), Richmond (RIC1) and Kansas City (MKC1) with Premium-CPU variants only; Memphis (MEM1) is listed with v5 and GPU Droplets only, and AMS2 and SFO1 are legacy with creation restricted. Google's footprint is wider and reaches Texas, the Mountain West and Los Angeles; DigitalOcean's reaches the Southeast, the Mid-Atlantic and Kansas City, where Google has no region. Prices are the same in every DigitalOcean region and differ by region on Google.
The platform around the server
The reason to pay either company more than a plain VPS host is what sits next to the server. DigitalOcean's platform is a short list a small team can hold in its head: managed databases (PostgreSQL, MySQL, Redis, MongoDB and others), a managed Kubernetes service, App Platform, Spaces object storage with a CDN, load balancers, volumes, VPCs and an API with official clients. Google Cloud's is the whole of Google Cloud: Cloud SQL and Spanner, GKE, Cloud Run and Cloud Functions, Cloud Storage, BigQuery, Pub/Sub, Vertex AI, IAM and a network that reaches every region. This page has evaluated neither platform; it notes that DigitalOcean's is smaller and priced on the same plan model as its Droplets, and Google's is larger and metered like its VMs, and that a developer who needs one specific Google service will find the choice made for them.
Who should pick which
- A known monthly bill for a server plus a database: DigitalOcean.
- A free machine to keep, or a $300 quarter to learn cloud infrastructure: Google Cloud.
- Texas, Utah, Nevada, Los Angeles, Ohio or Virginia: Google Cloud. Atlanta, Richmond or Kansas City: DigitalOcean.
- An application that uses BigQuery, Vertex, Pub/Sub or GKE: Google Cloud, because the server should be next to them.
- A container deployed without a VM: App Platform or Cloud Run, on the model you already pay for.
- A standalone server: neither; the fixed-price hosts are cheaper with transfer included.
What this page could not verify
- Google Cloud machine-type, disk and snapshot prices: rendered by script on Google's pages; only the Free Tier quantities and billing rules are quoted.
- App Platform and Cloud Run prices: on product pages not read for this version.
- DigitalOcean Premium CPU and v5 Droplet prices: on tabs and in the configurator, not in the default table.
- Which Droplet sizes are offered in ATL1, RIC1 and MKC1 beyond the Premium-CPU restriction: the product matrix on the regional availability page is the reference.
- Any performance figure: none published by either for these sizes, none measured here.