Spectres of different Species wait on a pier beside open bottles while a crane lowers another bottle toward the sea.

Pricing · Compute Nodes and Deployments

Compute for apps and AI

Choose CPU Compute Node resources by physical core count, RAM, and SSD, then account for Deployment snapshots separately.

Explore Prisman

Node capacity and Deployment storage are billed separately

A Node provides the capacity to run your workload. A Deployment is a stored snapshot, ready to run on a compatible Node. Its occupied storage is billed separately from the disk assigned to the Node.

Prisman pricing in practice

Choose resources for the workload

Start with the application’s compute needs and the snapshots you will keep ready to deploy.

Choose the Compute Node

Choose physical cores, RAM, and Node SSD according to the workload.

Choose the location and period

Choose the Node’s location when you contract it, then compare hourly and monthly compute rates.

Account for stored Deployments

Add the occupied GB of each snapshot to the storage total for Deployments.

Pricing structure

What goes into the compute price

Compute Node resources, stored snapshots, and the selected location all contribute to the compute cost. Model inference is a separate service with its own terms.

Node capacity

Each Compute Node profile provides physical cores, RAM, and SSD for the operating system, container images, and temporary work. Persistent application data is accounted for separately.

Monthly compute

A 730-hour month applies a 20% discount to cores and RAM. Node SSD remains a GB-month charge.

Deployment snapshots

One Deployment equals one snapshot. Its occupied storage costs $0.56 per GB per month, separately from the Node disk.

Inference and memberships

Prisman supplies inference services. Dumpla sells the memberships used for direct Dumpla participation; model inference and API consumption keep their respective service conditions.

Compute services

Compute and AI services

Explore Prisman services for container applications, AI workloads and inference.

Node compute

Container Apps

Package and operate applications as containers on Prisman Compute Nodes.

AI workloads

AI Agents

Run AI Agents alongside the compute and inference workloads they use.

Isolated environments

Sandbox

Run workloads inside an isolated environment.

AI compute

Inference Nodes

Serve model inference through Compute Nodes sized for the selected AI workload.

AI connectivity

AI Gateway

Connect AI applications with the model inference services they use.

Compute resource rates

Understand the unit prices

Compute Nodes combine physical cores, RAM, and Node SSD. Deployment snapshots add their own occupied-storage charge.

Physical core

$0.0980 / core-hour

Processing capacity for the Node’s workload.

Monthly: $57.232 / core
730 hours with the 20% monthly compute discount

RAM

$0.0084 / GB-hour

Memory assigned to the Node.

Monthly: $4.9056 / GB
730 hours with the 20% monthly compute discount

Node SSD

$0.42 / GB-month

Node disk for the system, container images, and temporary work.

Additional Node SSD uses this monthly unit rate
Application data storage has a separate scope

Deployment snapshot

$0.56 / GB-month

Each Deployment is a snapshot billed by the GB it occupies.

Stored separately from the Node’s assigned disk
Snapshot storage does not receive the compute discount

USD base rates before any location adjustment. Monthly compute receives a 20% discount; Node SSD and snapshots remain GB-month charges.

CPU Compute Nodes

Compare complete Node profiles

Each profile combines physical cores, RAM, and assigned SSD. Choose the location when contracting the Node.