Snowflake credit calculator: compare Snowflake, BigQuery and Databricks SQL cost

How it is calculated

h_idle   = r × max(s, 1) ÷ 60
h_billed = min(24, a + h_idle)        (24 when auto-suspend is off)

a = active hours per day, r = resumes per day, s = auto-suspend minutes

Credits = h_billed × c × n
C_SF    = Credits × price_per_credit + TB_stored × storage_rate

c = credits per hour for the size (XS 1, S 2, M 4, L 8, XL 16, 2XL 32, 3XL 64, 4XL 128), n = working days

TB_month = tb_scanned_per_day × n
Compute  = max(TB_month − free_TB, 0) × usd_per_TB
Storage  = active GiB × active rate + long-term GiB × long-term rate

Slot-hours = avg_slots × 24 × n
Compute    = Slot-hours × slot_hour_rate × (1 − commitment_discount)

DBU     = dbu_per_hour(size) × h_billed × n
Compute = DBU × (usd_per_DBU + cloud_usd_per_DBU_hour)

The cloud term is 0 on serverless; classic and pro add an estimated cloud compute cost.

saved credits = (h_billed − h_billed with s = 1) × c × n
saved USD     = saved credits × price_per_credit
saving %      = saved USD ÷ Snowflake compute

scan* = (Snowflake compute ÷ usd_per_TB + free_TB) ÷ n   TB per day

Below scan*, BigQuery on-demand costs less than Snowflake compute; above it, Snowflake costs less.

Assumptions

• The comparison assumes the stated warehouse can serve the stated scan volume. Warehouse size and bytes scanned are different cost drivers, and no size is checked against the volume.
• Idle time is an upper bound: every resume is assumed to be followed by a full auto-suspend window. Real idle time is lower when queries arrive before the window ends.
• Billed hours are capped at 24 a day. Auto-suspend set to 0 is treated as 1 minute, because Snowflake bills at least 60 seconds per resume. The never-suspend switch bills all 24 hours.
• Databricks SQL uses the same billed hours as Snowflake, with its auto stop taken to equal the auto-suspend setting. Its warehouse size is the nearest one on the same doubling ladder (a modelled pairing, not a performance equivalence).
• Snowflake prices are On Demand list prices for standard (Gen1) warehouses. Gen2 and Snowpark-optimized warehouses, multi-cluster scaling, serverless features, data transfer and cloud services are not modelled.
• BigQuery is priced in binary units; one TB here is one TiB (1,024 GiB). Storage uses logical bytes. The 10 MB per-query minimum and the free 10 GiB of storage are ignored. BigQuery editions price the average slots around the clock; autoscaling spikes are not modelled.
• Databricks cloud compute for classic and pro warehouses is an estimate; Databricks storage is your own cloud storage at an assumed object-store rate.
• Prices are USD list prices. Negotiated discounts, free credits and taxes are not applied; enter your Snowflake contract price to override the list price.

The data behind it

Three ways to reduce this number

1. Set auto-suspend to 60 seconds so idle tails stop billing.
2. Partition and cluster tables, and select only the columns you need, so each query scans fewer bytes.
3. Commit capacity (a Snowflake capacity contract or a BigQuery slot commitment) once usage is predictable.

Use it from code

The same engine is available as a REST endpoint and as an MCP tool named calc_data_warehouse_credit_estimator. Your scenario as an API call:

curl -X POST 'https://api.x12i.com/v1/calc/data-warehouse-credit-estimator' \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
  "input": {
    "platforms": [
      "snowflake",
      "bigquery_on_demand",
      "databricks_sql"
    ],
    "warehouse_size": "M",
    "snowflake_edition": "enterprise",
    "cloud_region": "aws-us-east-1",
    "active_hours_per_day": 8,
    "resumes_per_day": 12,
    "auto_suspend_minutes": 10,
    "auto_suspend_never": false,
    "working_days_per_month": 30,
    "tb_scanned_per_day": 4,
    "storage_tb": 30,
    "long_term_storage_pct": 50,
    "bq_edition": "enterprise",
    "bq_commitment": "none",
    "databricks_tier": "serverless",
    "data_warehouse_stack": "none"
  },
  "options": {
    "explain": false
  }
}'

Copy as API call Schema: GET /v1/calc/data-warehouse-credit-estimator/schema

Frequently asked questions

Related calculators