Skip to the answer
Proxy Costsobserved 2026-07-29

Term

Defined by what it does to a bill, with a real plan from the dataset where the term changes the price.

Minimum checkout cost: the first bill is not the advertised unit rate

Minimum checkout cost is the smallest total a provider will accept for a purchasable proxy configuration. It changes the bill by replacing a tempting unit rate with the amount that must actually leave the buyer’s account. A rate per GB or per IP is useful only beside the quantity attached to it. If the provider sells that rate inside a package, buying less traffic does not shrink the checkout total unless a smaller package is also published.

A real rate with a real checkout

Oxylabs names its plan Shared (pay per GB) in the dataset. The first published datacenter package contains 20 GB at $0.59 per GB and has an $11.80 monthly checkout. The same ladder continues through larger packages, but the first row is enough to show why the unit rate is not a receipt. Oxylabs shared datacenter pricing, retrieved 2026-07-25T15:31:32Z

A buyer who needs only 1 GB cannot take the displayed $0.59 rate and claim a $0.59 bill. The published configuration that covers the request still costs $11.80 and leaves 19 GB unused. For that buyer, the checkout cost is twenty times the displayed unit rate because the purchase quantity is twenty times the requested quantity. Oxylabs shared datacenter price ladder, retrieved 2026-07-25T15:31:32Z

The distinction also prevents a subtler pricing error. Oxylabs lists a 50 GB package at $0.57 per GB and $28.50 monthly. That lower unit rate belongs to the 50 GB package. It cannot be multiplied by 20 GB to create a cheaper configuration that the provider does not sell. Each package keeps its own quantity, rate, and checkout total. Oxylabs published datacenter packages, retrieved 2026-07-25T15:31:32Z

How often providers publish the checkout

In the current snapshot, 41 of 47 providers publish min_checkout_usd on at least one plan. The field contains a real total, including zero only where the provider explicitly publishes a free package. The other 6 providers have no published checkout total in the collected plan rows. Their null values do not mean that signup or usage is free. Current 710-plan dataset snapshot

A comparison should therefore start with the requirement, find every published package that covers it, and compare checkout totals. Only after that should it show a per-unit figure. Record the package quantity, billing period, expiry, taxes, and whether the balance can roll over. These terms decide whether unused capacity still has value.

When a provider shows only a starting rate without the purchasable quantity or total, the checkout remains not published. Multiplying that rate by a convenient quantity would manufacture an offer. A defensible result either names the complete package and its total or excludes the row from a cheapest-checkout claim.