Price lists for cloud data warehouses can be misleading. They don't clearly convey how billing works for compute resources, which is vital for anyone managing analytical workloads. The article breaks down how five major cloud data warehouses—Snowflake, Databricks (SQL Serverless), ClickHouse Cloud, Google BigQuery, and Amazon Redshift Serverless—translate queries into costs. Understanding this helps in comparing the actual performance per dollar across these platforms. The author emphasizes that each platform uses different compute billing units, like credits or compute units, which complicates direct comparisons. Factors like how query runtimes scale with data size and variations in compute consumption for the same query on different engines further muddy the waters. The article introduces the Bench2Cost framework, designed to standardize cost comparisons by linking benchmark results to real pricing models. This framework allows users to see how much compute a query actually costs in dollars, giving a clearer picture of expenses across platforms. In the section on Snowflake, the article details its compute model, which relies on fixed-size warehouses billed by the hour. Prices vary by edition, from $2.00 per credit for Standard to $4.00 for Business Critical. Snowflake can scale vertically by resizing warehouses or horizontally by adding clusters during high demand, both affecting costs. It also offers a Query Acceleration Service for additional fees, aimed at optimizing performance for demanding queries. Understanding these nuances is key to managing costs effectively when using cloud data warehouses.