Buying snapshot
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Decision point | What SingleStore states |
|---|---|
Billing basis | Cost is determined by actual usage: workspace compute billed in credits per hour, plus storage per GB. |
Cloud providers | AWS, Azure or Google Cloud; you also pick a region. |
Data ingress | SingleStore does not charge for data ingress, but your cloud provider may. |
Estimates and discounts | Listed prices are estimates. Sales gives more accurate pricing and volume-based discounts, and a commitment pricing model is offered when an application moves to production. |
Which edition fits
Shared if you are evaluating, developing or testing outside production; it costs nothing.
Standard for general-purpose production. Plan a two-zone (2 AZ) deployment if you need the 99.99% SLA.
Enterprise for customer-facing workloads that need recovery and security controls. Point-in-time recovery (PITR) is an Enterprise feature in the Helios docs.
BYOC Standard if managed Helios must run inside your own VPC.
Editions and published prices
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Edition | Published price | Intended use and key features |
|---|---|---|
Shared (Managed) | 1 Free Workspace | For evaluation, development and non-production testing. Shared workspace, all SQL and programmability features, monitoring, self-support. |
Standard (Managed) | Starts at $0.99/hr | For general-purpose production workloads. Read-replicas, all SQL and programmability features, database branching, monitoring (30-day), 99.9% SLA (1 AZ) or 99.99% SLA (2 AZ), Standard Support. |
Enterprise (Managed) | Starts at $1.49/hr | Enhanced DR and security for customer-facing workloads. Everything in Standard, plus online PITR, Smart DR, audit logging and CMEK; 99.9% SLA (1 AZ) or 99.99% SLA (2 AZ). |
BYOC Standard | Contact us for pricing | Fully managed Helios Cloud within your own VPC. Read-replicas, all SQL and programmability features, monitoring (30-day), Standard Support. |
How the bill is built
The "starts at" rates are entry points, not the rate for every workspace. SingleStore's estimator asks for edition, cloud provider, region, workspace size (memory), hours per day and days per month. It returns a monthly credit cost for compute and a monthly storage cost.
Compute is metered in credits per workspace hour, and the credit rate rises with size. In the vendor's examples, S00 uses 0.25 credits per hour, S4 uses 4 and S8 uses 8, each at $3.96 per credit. Storage is billed per GB on the month's average stored data: $0.023 per GB in the Tier-1 example and $0.025 per GB in the Tier-2 example.
SingleStore's worked examples
These are the vendor's own bills, with its assumptions. Examples #1 and #2 cover database compute and storage only; data movement or ingestion with third-party tools is not included. Example #3 is a separate bill for the optional Flow service.
Scroll sideways to see all columns. With a keyboard, focus this table and use the arrow keys.
| Example | Assumptions | Vendor's stated bill |
|---|---|---|
#1 Learn, develop and test | S00 Standard in a Tier-1 region, 8 hours x 20 days; 100 GB stored for 15 days and 500 GB for the other 15; 500 GB ingested. | Compute: 40 credits x $3.96/credit = $158.40. Storage: 300 GB average x $0.023/GB = $6.9. Ingress: $0. |
#2 Multi-tenant production plus dev/test | Tier-2 region. S4 (free tenants) and S8 (paid tenants) up all month; S00 dev on 8 hours x 20 days; S8 test on 4 hours x 5 days. | Compute: (2,880 + 5,760 + 40 + 160) credits x $3.96/credit = $35,006.40. Storage: 22,995 GB x $0.025/GB = $574.9. Ingress: $0. |
#3 CDC replication with Flow (optional) | Flow sized at F-3, running 24x7; about 10 TB processed, including the initial load and ongoing CDC updates. | Flow compute: 0.927 credits/hr x 730 hrs at $3.96/credit = $2,679.77. Data processed: 10,240 GB at $0.02/GB = $204.80. |
What drives the bill
Workspace size. Credits per hour grow with size, and cluster sizes go up to S-384 (3,072 vCPU, 24,576 GB memory).
Running hours. A workspace bills for each hour it is up. Example #2 bills production workspaces that are up throughout the month for 24 hours x 30 days.
Stored data over time. Storage is averaged across the month, so data that grows mid-month counts only for the days it is stored.
Costs that are easy to miss
Recovery snapshots. The Helios docs say data kept in the storage tier may exceed the size of your active databases, because snapshots are retained for recovery.
Suspend and resume. The docs say a suspended cluster is torn down completely, and its local cache, including the plan cache, is lost; a resumed cluster starts cold.
Ingestion. Flow is optional and billed separately in credits plus per GB processed (example #3). Third-party ingestion tools are outside SingleStore's estimates.
Our take
If you can start on the free Shared workspace and move to production at the size and hours you need, SingleStore's usage model is easy to test before you pay. If you plan a production workspace that runs all month, price your real size in the estimator and ask sales about commitment pricing before you sign.