Price a Redshift cluster across node hours, managed storage, concurrency scaling above the free credits and Spectrum scans.
Node-hours dominate, but the two components people forget are concurrency scaling and Spectrum. Redshift earns one free hour of concurrency scaling for every 24 hours the main cluster runs, so a cluster that is up all month accrues about 30 free hours and only the excess is billed at the same per-node rate — and Spectrum queries are priced per TB scanned exactly like Athena, entirely separately from the cluster. Every rate is an editable input with a realistic us-east-1 default, so confirm current Redshift prices with AWS for your region. A pause-and-resume schedule cuts node-hours but also cuts the concurrency-scaling credits it earns, so the cheapest configuration is rarely the one with the fewest running hours.
Redshift Cluster Cost
monthly = nodes × hourly × hours × (1 − reserved discount) + managed GB × rate + max(0, scaling hours − credits) × nodes × hourly + Spectrum TB × rate, where credits = 1 free hour per 24 hours of cluster uptime up to 30.
monthly = nodes × hourly × hours × (1 − reserved discount) + managed GB × rate + max(0, scaling hours − credits) × nodes × hourly + Spectrum TB × rate, where credits = 1 free hour per 24 hours of cluster uptime up to 30. Node-hours dominate, but the two components people forget are concurrency scaling and Spectrum. Redshift earns one free hour of concurrency scaling for every 24 hours the main cluster runs, so a cluster that is up all month accrues about 30 free hours and only the excess is billed at the same per-node rate — and Spectrum queries are priced per TB scanned exactly like Athena, entirely separately from the cluster. Every rate is an editable input with a realistic us-east-1 default, so confirm current Redshift prices with AWS for your region.
A pause-and-resume schedule cuts node-hours but also cuts the concurrency-scaling credits it earns, so the cheapest configuration is rarely the one with the fewest running hours.
This calculator takes 9 inputs: Nodes in the cluster, Node price, Cluster hours per month, Reserved node discount, RA3 managed storage, Managed storage price, Concurrency scaling used, Redshift Spectrum data scanned, Spectrum price. The pre-filled defaults are a realistic starting point — replace them with figures from your own environment for a result you can act on.
The defaults are realistic us-east-1 on-demand figures for RA3, but Redshift pricing varies by region and node family and Serverless is billed per RPU-hour instead. Confirm against the AWS Redshift pricing page for your region.