SPG99 pricing

SPG99 is a usage-based PostgreSQL service for dev/test, preview/staging, internal services, and other uneven-load scenarios. You pay for actual database runtime rather than 24/7 idle time.

Price display

All prices below are listed in RUB incl. VAT.

Profile availability

L1 and L2 are available now. L3-L5 are extended profiles available on request.

Profiles

Compute is billed at the rate of the selected profile. Below you can see the base active-hour rate and the reference totals for 100 and 400 active hours per database.

L1Available now

2 vCPU / 4 GB RAM

9.50 ₽ / active hour
100 active hours950 ₽
400 active hours3,800 ₽
L2Available now

4 vCPU / 8 GB RAM

19.00 ₽ / active hour
100 active hours1,900 ₽
400 active hours7,600 ₽
L3On request

8 vCPU / 16 GB RAM

38.00 ₽ / active hour
100 active hours3,800 ₽
400 active hours15,200 ₽
L4On request

16 vCPU / 32 GB RAM

76.00 ₽ / active hour
100 active hours7,600 ₽
400 active hours30,400 ₽
L5On request

32 vCPU / 64 GB RAM

152.00 ₽ / active hour
100 active hours15,200 ₽
400 active hours60,800 ₽
Profile availability

The public pricing model already exposes the full L1-L5 line. Today the active runtime range is L1 and L2, while larger profiles are enabled on request for expanded scenarios.

What goes into the bill

The monthly bill is composed of a few transparent line items: compute for the selected profile, storage, egress, and storage operations.

Compute

Database compute resources billed at the rate of the selected profile.

Profile rate
Storage

Average occupied database volume during the month.

2.30 ₽ / GB per month
Egress

Actual outgoing traffic during the month.

1.70 ₽ / GB
GET / HEAD / OPTIONS

Storage operations.

0.43 ₽ / 10,000 ops
What are active hours

Active hours are the actual time when the database is running and ready to accept connections.

SPG99 measures database runtime rather than CPU utilization. If the database is idle and compute is stopped, no compute hours are billed.

How the bill is calculated

Monthly bill = total compute across all databases + storage + egress + storage operations.

Compute
profile rate × active hours
Storage
average occupied volume for the month × 2.30 ₽
Egress
actual outgoing traffic × 1.70 ₽
GET / HEAD / OPTIONS
number of ops / 10,000 × 0.43 ₽

If you have multiple databases, each database is calculated first and then all line items are summed into one bill.

Calculation examples

Three baseline scenarios are shown below: one small database, one larger database with operations, and several identical dev/test environments.

Example 1

One L1 database, 100 active hours, 30 GB storage, 20 GB egress.

  • Compute: 100 × 9.50 = 950 ₽
  • Storage: 30 × 2.30 = 69 ₽
  • Egress: 20 × 1.70 = 34 ₽
Total: 1,053 ₽
Example 2

One L2 database, 400 active hours, 50 GB storage, 150 GB egress, and 300,000 GET / HEAD / OPTIONS.

  • Compute: 400 × 19.00 = 7,600 ₽
  • Storage: 50 × 2.30 = 115 ₽
  • Egress: 150 × 1.70 = 255 ₽
  • Operations: 300,000 / 10,000 × 0.43 = 12.90 ₽
Total: 7,982.90 ₽
Example 3

Three L1 databases, each active for 160 hours per month, average storage 20 GB per database, total egress 90 GB.

  • Compute: 3 × 160 × 9.50 = 4,560 ₽
  • Storage: 3 × 20 × 2.30 = 138 ₽
  • Egress: 90 × 1.70 = 153 ₽
Total: 4,851 ₽
When the model works best

When the model works best

SPG99 is especially cost-effective where the database is not needed around the clock: dev/test, preview/staging, internal services, background jobs, and temporary environments.

dev/testpreview/staginginternal servicesbackground jobstemporary environments

If the database runs almost continuously throughout the month, the economic effect compared with classic managed PostgreSQL will be lower.

Calculator

Cost and savings calculator

The calculator estimates the SPG99 monthly bill from your actual parameters and shows a managed HA benchmark beside it. The comparison uses a managed HA baseline rather than a simplified single-node baseline because SPG99 operates at a higher reliability class than a simple primary + replica topology.

L12 vCPU / 4 GB RAMAvailable now
How to enter time

You can enter either active hours or idle hours per database for the benchmark month.

The estimate uses
100 h
per database · total active hours: 100 h
630 h idle / 730 h benchmark month

Monthly estimate

The SPG99 bill is always calculated only from your actual inputs. The managed block below is an aggregated average HA benchmark based on managed PostgreSQL pricing in Russia for a comparable profile.

SPG99 bill
1,053 ₽
Compute: 950 ₽
Managed PostgreSQL HA benchmark
13,973 ₽
Comparable managed HA profile; aggregated average from managed PostgreSQL pricing in Russia, 730 h/mo
Savings
12,920 ₽
Managed compute benchmark: 13,870 ₽
Savings, %
92.46%
Total active hours: 100 h

Breakdown

Storage, egress, and operations stay at the same monthly volumes. Only the compute model changes in the benchmark.

Compute
950 ₽
Storage
69 ₽
Egress
34 ₽
GET / HEAD / OPTIONS
0 ₽
SPG99 bill
1,053 ₽
Managed compute benchmark
13,870 ₽
Managed total benchmark
13,973 ₽
How to read the comparison
  • The comparison uses a managed HA benchmark rather than a single-node baseline because SPG99 sits in a higher reliability class than a simple primary + replica topology.
  • The managed price in the calculator is an aggregated average benchmark based on managed PostgreSQL pricing in Russia for a comparable HA profile.
  • Storage, egress, and operations are carried over with the same monthly volumes so the comparison does not understate or overstate savings through other line items.
  • If you enter idle hours, the calculator automatically converts them into active hours for the estimate.
Next step

Validate the model on your actual workloads

If you want to estimate your team scenario quickly or run a short pilot, leave a request. We will walk through the numbers, answer the questions, and help assemble a comparable setup.