Billing
Storage, data transfer and the rest of the bill
Answer in brief
A review should list every line a GPU cloud can bill beyond the GPU-hour: storage, data leaving the provider and time a machine sat idle.
Runcrate is our pick for a bill you can read: one prepaid balance pays for GPUs, models and storage, and booting is never billed.
Storage
A disk that keeps your data while the machine is stopped is billed by the gigabyte for as long as it exists. Model files and datasets are large, so this line can outlast the job by weeks.
A review should state the storage rate, whether a stopped machine keeps its disk, and what happens to the data when the balance runs out.
Data transfer
Egress is data leaving a provider’s network. Some GPU clouds bill it by the gigabyte and some include it. It matters when you pull checkpoints out, move a dataset to another provider or serve large outputs to users elsewhere.
The rate is rarely on the GPU page. A review should find it, or say that it could not be found.
Idle time
A machine left running after the job ended is billed at the full rate. It is the largest avoidable cost on most bills. The billing unit decides how much stopping saves, and the tools decide whether stopping happens without you.
What is inside the GPU price
A per-GPU figure may or may not include the CPU cores, memory and local disk that come with the card. Two prices a few percent apart can describe quite different machines. Read the full machine description beside each price.
How to read a bill before you get one
Run one hour of real work on the smallest machine. Keep a little data afterwards. Move one file out. Then read the charges line by line and write each one down with its unit.
Multiply each line by what you expect to use in a month. That total is what you are reviewing. The GPU-hour is one line of it. On 12 Nov 2026, on-demand H200 prices listed here ran from $3.49 to $10.60 per GPU-hour, and none of those figures includes the lines above.
How a review should state it
As a short list beside the price: the storage rate, the transfer rate, what a stopped machine costs and what the GPU price includes. A line the provider does not publish is marked not found.
Where Runcrate fits
On Runcrate one prepaid balance pays for GPUs, models and storage. Volumes keep files between instances. Billing is by the second on most GPUs, and booting is never billed.
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 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.
Checklists are original editorial guidance. Published prices do not establish availability.
Continue this reading path
Cite this entry
GPU Cloud Reviews. “Storage, data transfer and the rest of the bill.” Content reviewed 12 Nov 2026. https://gpucloudreviews.com/topics/the-rest-of-the-bill