How This Snowflake Cost Estimator Works

Every number you see in the results panel traces back to a formula. Here is what the calculator is doing behind the scenes for each cost component.

Compute Cost

Compute is almost always the largest portion of your Snowflake bill. Snowflake bills credits per second with a minimum charge of 60 seconds every time a warehouse starts.

The Formula:

Compute Cost = Credits per Hour × Hours per Day × Days per Week × 4.33 × Credit Price

4.33 is the average number of weeks in a month. A query running 15 minutes on a Small warehouse (2 credits per hour) at $2.00 per credit costs $1. Scale that to a full month and the calculator multiplies your daily active hours by 4.33 weeks to get monthly compute spend.

How Credit Price Is Determined

Credit price depends on three things: your cloud provider (AWS, Azure, or GCP), the region you deploy in, and the edition you are on. US East on AWS Standard is $2.00 per credit. Enterprise in EU Frankfurt on AWS is $3.90 per credit. The calculator applies the correct rate automatically once you select your configuration. For a full edition and pricing breakdown, read All You Need to Know About Snowflake Pricing.

Storage Cost

Storage is billed as a flat rate per terabyte per month. Snowflake compresses your data automatically, typically achieving 3x to 5x reduction on raw data, so your billed storage volume is usually much lower than your raw data size.

The Formula:

Storage Cost = Storage Volume (TB) × Regional Rate ($/TB per month)

Most US regions cost approximately $23 per TB per month. EU and APAC regions are slightly higher.

Time Travel and Fail Safe Storage (Advanced Mode)

Time Travel lets you query your data as it existed at any point in the past. Extending retention past 1 day increases your storage bill because Snowflake keeps pre-modified versions of your data alongside the current version.

The Formulas:

Time Travel Cost = Storage (TB) × Data Change Rate (%) × (Retention Days ÷ 30) × Storage Rate

Fail Safe Cost = Storage (TB) × Data Change Rate (%) × (Fail Safe Days ÷ 30) × Storage Rate

For a detailed guide on what Time Travel actually costs at scale, read Snowflake Time Travel Guide for Cost Implication.

ETL Compute (Advanced Mode)

ETL pipelines run on their own dedicated warehouse. The 60-second minimum billing rule hits hardest here. A pipeline that completes in 10 seconds costs the same as one that runs for a full minute.

The Formula:

ETL Cost = Credits per Hour × (Runs per Day × 30) × Max(Duration mins, 1) ÷ 60 × Credit Price

Query Compute (Advanced Mode)

Analytics queries from active users and BI tools run on a separate query warehouse. The calculator models both human-triggered queries and automated BI dashboard refreshes.

The Formulas:

Queries per Month = (Users × Queries per User per Day + BI Tools × 50) × 30

Query Hours per Month = Queries per Month × Max(Avg Query Duration minutes, 1) ÷ 60

Query Cost = Credits per Hour × Query Hours per Month × Credit Price

The BI Tools multiplier of 50 accounts for automated refreshes that run independently of human users. Query result caching can bring this cost close to zero for repeated queries. Read Snowflake Query Caching and Performance to understand when cache applies.

Cloud Services (Advanced Mode)

Snowflake does not bill cloud services unless they exceed 10% of your daily virtual warehouse compute. If your account consumes 150 credits in a day, you get 15 cloud services credits at no charge. Only usage above that threshold is billed. For the vast majority of accounts this rounds to zero.

Data Transfer (Advanced Mode)

Snowflake only charges for data leaving the platform. Loading data in is free.

The Formula:

Transfer Cost = Volume (TB) × 1000 × Rate ($/GB)

Same cloud, same region costs $0 per GB. Same cloud, different region costs approximately $0.02 per GB. A different cloud provider or internet egress costs approximately $0.09 per GB.

The Estimate Range

The calculator outputs a low to high range rather than a single figure. This range reflects real world variance from caching behavior, auto-suspend settings, and warehouse scaling.

The Formulas:

Low Estimate = Base Total × 0.85  |  High Estimate = Base Total × 1.25

A warehouse left running past its configured hours due to a long query, or one with auto-suspend disabled, will land closer to the high end. Setting auto-suspend to 60 seconds is the single fastest way to tighten this range. Read Snowflake Auto Suspend Cost Management for configuration guidance.

After You Get Your Snowflake Cost Estimate

If your estimate is higher than expected, three settings move the number the most: enabling auto-suspend at 60 seconds, dropping your warehouse one size tier, and separating ETL and query workloads onto dedicated warehouses. For a full optimization playbook read How to Optimize Snowflake Cost and The Complete Guide to Snowflake Cost Optimization.

Running this estimate ahead of a migration? The calculator gives you steady-state monthly spend. For modeling first-year costs including migration overhead and data growth, read Snowflake Cost Forecasting: How to Model Spend Before Migration. Finance or procurement reviewing this number? Read How CFOs Should Think About Snowflake Pricing for guidance on on-demand versus capacity commitment trade-offs.