Google CloudBigQueryAnalytics

`google_bigquery_reservation` cost estimation

Dedicated BigQuery compute capacity (slots) for predictable query pricing. Priced per slot-hour by edition, billed for the reserved slot count.

A google_bigquery_reservation buys dedicated BigQuery query capacity measured in slots, as an alternative to on-demand per-TB-scanned pricing. For teams running heavy, steady query workloads, reservations make spend predictable: you pay for the slots you reserve rather than per byte each query scans.

Slots bill per slot-hour by edition. Standard edition is roughly $0.04/slot-hour, Enterprise ~$0.06, and Enterprise Plus ~$0.10, with the higher editions adding features like multi-region replication and longer commitments. A reservation of 100 Enterprise slots runs about $0.06 x 100 x 730 = ~$4,380/month. Slots can be bought as flex (per-second, no commitment), monthly, or annual commitments at decreasing rates.

c3x reads the slot_capacity and prices slot-hours at 730 hours/month at the edition rate. The break-even against on-demand depends on how many terabytes you'd otherwise scan: above roughly the point where on-demand would cost more than the reservation, slots win. c3x lets you compare both before committing.

Terraform example

A minimal but realistic configuration that C3X can estimate.

resource "google_bigquery_reservation" "analytics" {
  name           = "analytics"
  location       = "US"
  slot_capacity  = 100
  edition        = "ENTERPRISE"

  autoscale {
    max_slots = 200
  }
}

Pricing dimensions

What you actually pay for when you provision google_bigquery_reservation.

Dimension	Unit	What's being charged
Reserved slots	per slot-hour	Per-slot hourly rate by edition, billed for the reserved slot_capacity. c3x assumes 730 hours/month. $0.04 Standard, $0.06 Enterprise, $0.10 Enterprise Plus
Autoscale slots	per slot-hour	Slots added by autoscaling above the baseline, billed per second only while active. Usage-based.
Commitment discount	monthly / annual	1-year and 3-year commitments lower the per-slot rate versus flex (no-commitment) slots.

Sample C3X output

Example output from c3x estimate (100 Enterprise slots):

google_bigquery_reservation.analytics
└─ Reserved slots (Enterprise x 100)   73,000  slot-hours    $4,380.00

OVERALL TOTAL                                                $4,380.00

Optimization tips

Common ways to reduce google_bigquery_reservation cost without changing the workload.

Compare reservations against on-demand first

Break-even dependent

On-demand BigQuery charges per TB scanned. Reservations win only above a query volume where on-demand would exceed the slot cost. Estimate both for your real scan volume before reserving.

Use baseline + autoscale instead of a big fixed reservation

Peak headroom

Set a small committed baseline for steady load and let autoscale handle spikes per-second. This avoids paying for peak capacity 24/7.

Commit for 1 or 3 years on stable baselines

Commitment discount

Committed slots cost less per slot-hour than flex. For a predictable analytics baseline, an annual commitment trims the rate meaningfully.

FAQ

How does c3x estimate a BigQuery reservation?

It reads slot_capacity and edition and prices the reserved slot-hours at 730 hours/month. Autoscale slots are usage-based (billed per-second while active) and supplied in c3x-usage.yml.

When do slot reservations beat on-demand pricing?

On-demand bills ~$6.25 per TB scanned. Reservations bill flat per slot-hour. Above the query volume where on-demand cost would exceed the reservation, slots are cheaper and far more predictable. Below it, on-demand wins.

What's the difference between the editions?

Standard (~$0.04/slot-hr) is the entry tier; Enterprise (~$0.06) adds features and commitment options; Enterprise Plus (~$0.10) adds multi-region and advanced governance. The edition sets the per-slot rate.

Does storage cost extra?

Yes. Reservations cover query compute only. BigQuery storage (active and long-term) bills separately on the dataset, which c3x prices as google_bigquery_dataset.

What are flex slots?

Flex slots are reserved per-second with no commitment, ideal for short bursts of heavy querying. Committed (monthly/annual) slots cost less per hour but lock in the capacity.