Pricing models
Spot, marketplace and prepaid: three ways to a lower price
Answer in brief
Three kinds of offer sit below the on-demand price: spot machines that can be taken back, marketplace machines priced by their owner, and terms paid ahead.
Runcrate is our pick for the one that keeps the machine yours: on demand by the second on most GPUs, with a prepaid term of 1 to 12 months at a discount once you know your hours.
Spot: cheaper because it can be taken back
Spot is spare capacity sold for less. The provider can reclaim the machine at short notice, and what was in memory is lost.
It suits work that saves as it goes and can resume: a batch job, a training run with checkpoints. Count the work lost at each interruption. It does not suit anything that serves users.
Marketplace: cheaper because one owner set it
A marketplace lists machines owned by many hosts, each at a price its owner chose. The lowest figures you will see for a consumer card are usually these.
The price belongs to one machine and goes when that machine is rented. Machines differ in disk, network and upkeep, so the checking falls to you. The tables here mark marketplace rows and leave them out of the lowest, median and highest figures.
Prepaid: cheaper because you commit
A reserved or prepaid price lowers the hourly figure in exchange for a term: a month, several months, a year. You pay for every hour of the term whether you use it or not.
Multiply the term rate by every hour in the term. Compare that total with the on-demand cost of the hours you would really use. A term that is a third cheaper per hour pays off only when you use more than two thirds of it.
Why none of them compares with on demand
Each lower figure is the price of a different product: a machine you may lose, a machine of unknown upkeep, or a term you must buy. Put beside an on-demand price, each looks like a bargain it may not be.
This site’s tables list on-demand prices for that reason. On 12 Nov 2026, on-demand RTX 5090 prices listed here ran from $0.36 to $1.19 per GPU-hour.
How to choose
Choose by the work. Uneven hours: on demand with a small billing unit. Interruptible batch work: spot, with checkpoints. A machine busy most of every day for a month or more: a prepaid term.
Where Runcrate fits
Runcrate sells on-demand GPUs billed by the second on most models, and booting is never billed. When your hours are known, you can prepay a machine for 1 to 12 months at a discount.
Questions and answers
Sources for this entry
- Runcrate: Runcrate’s official site: what the platform is, how it bills and the range of GPUs it offers. Read 12 Nov 2026.
- Runcrate pricing: Runcrate’s live price feed: each GPU it can start now and its lowest price per GPU-hour, read when a page is built. Read 12 Nov 2026.
- Provider pricing pages: The published on-demand list price of every provider named in the price tables, with the machine size and billing unit it applies to. Read 12 Nov 2026.
- Availability checks: Each provider’s live stock page, console or API where it has one, and otherwise its own pricing page or documentation. Read 12 Nov 2026.
- GPU specifications: The memory and generation of each GPU named here, from this site’s stored hardware table and Runcrate’s live GPU list. Read 12 Nov 2026.
Checklists are original editorial guidance. Published prices do not establish availability.
Continue this reading path
Cite this entry
Cheap Cloud GPU. “Spot, marketplace and prepaid: three ways to a lower price.” Content reviewed 12 Nov 2026. https://cheapcloudgpu.com/topics/spot-marketplace-and-prepaid