Pricing & billing
Work is priced by the size of what you send. One credit is one US dollar, and there is no exchange rate to reason about.
The rates on this page are read from
GET /api/public/pricing when it loads, so they cannot fall behind what the
invoice will say. The endpoint is unauthenticated — you are welcome to read it directly.
Rates
| Engine | Per GB processed | |
|---|---|---|
| Voxelion Health | $0.015 / GB | Medical imaging. Near-identical slices matched by perceptual hash. |
| Voxelion Curate | $0.015 / GB | Product image archives and generative training corpora. Matched by perceptual hash. |
Size is a proxy for cost, not a measure of it: a 400-slice CT series and a folder of photographs of the same size do different amounts of work. It is the proxy you can verify from your own dataset without a rate card, which is what makes it the right one to bill on.
How credits work
- A run reserves its estimated cost before it starts. If you cannot afford
the estimate, it is refused with
402and nothing is processed. - The run executes.
- It settles the true cost. The difference between the reservation and the settlement returns to your balance.
Charges round to the nearest cent once per run, with a one-cent minimum on work of measurable size. Rounding happens once, at the end — not per file.
Top-up packs are convenient amounts, not discounts: paying $25 puts $25 of credit in the balance, the same as any other route in. There is exactly one answer to what a dollar of credit is worth.
What costs nothing
In-browser runs are free and always will be, because they do not touch our infrastructure. The console's local mode and the no-account demo hash images in your browser; nothing is uploaded and nothing is metered.
Server-side processing is what is billed, and it is billed from prepaid credit only — there is no invoice, no seat licence and no minimum commitment.
Enterprise
On-premise and private-cloud deployment, custom engine tuning and dedicated support are available. Talk to us.
Reference generated from live API (https://api.voxelion.ai) on 2026-08-30. The endpoint list, error codes and limits on this page are produced from the API's own route table — if an endpoint is not listed here, it is not enabled on production.