How Does Snowflake Compute Pricing Compare to BigQuery and Redshift in 2027?
The three warehouses meter compute in different units, so there is no single price to compare. Snowflake bills credits per hour by warehouse size, and a credit costs $2.00 on the Standard edition in AWS US East (N. Virginia), rising to $3.00 on Enterprise. BigQuery charges $6.25 per TiB scanned on demand (first 1 TiB each month free), or $0.04 to $0.10 per slot-hour on capacity editions. Redshift Serverless charges $0.375 per RPU-hour in US East (N. Virginia), and provisioned clusters bill by the node-hour.
All three bill per second after about a one-minute minimum, and all three stop charging for compute when idle in their serverless or auto-suspend modes. The real difference is what you pay for: Snowflake and Redshift charge for time a cluster is running, BigQuery on-demand charges for bytes read, and BigQuery capacity charges for slot time.
In the worked example below, 176 busy hours a month comes to $1,408 on a Snowflake Medium warehouse, $2,112 on a 32-RPU Redshift Serverless setup and $1,056 on 100 BigQuery Enterprise slots. Those blocks are not performance-equivalent, so treat them as price tags, not a benchmark.
How each platform bills compute
- Snowflake: credits per warehouse-hour. Each virtual warehouse consumes credits at a fixed hourly rate set by its size: 1 for X-Small, 2 for Small, 4 for Medium, 8 for Large, 16 for X-Large, doubling up to 512 for 6X-Large (Service Consumption Table, Table 1(a)). A suspended warehouse uses no credits; a started or resumed one is billed a minimum of one minute and then per second.
- BigQuery: bytes scanned, or slots over time. On-demand pricing charges for the bytes each query processes. Capacity pricing charges for slots (virtual CPUs) over time in the Standard, Enterprise or Enterprise Plus edition, with an autoscaler or optional one- and three-year commitments (BigQuery pricing).
- Redshift: node-hours or RPU-hours. Provisioned clusters bill per node-hour. Serverless meters Redshift Processing Units (RPUs), billed per second with a 60-second minimum and no charge for start-up time (Redshift pricing).
Two details matter early. First, Snowflake credits convert to dollars through your edition, cloud and region, so the same Medium warehouse costs more in Frankfurt than in Virginia. Second, BigQuery on-demand pricing follows data read, not run time, so a slow query and a fast one over the same bytes cost the same.
Snowflake compute pricing: credits, editions and regions
Table 2(a) of the Service Consumption Table (effective October 15, 2026) lists on-demand prices per credit. For AWS US East (Northern Virginia) it is:
- Standard$2.00 per credit
- Enterprise$3.00 per credit
- Business Critical$4.00 per credit
- Virtual Private Snowflake (VPS)$6.00 per credit
Region moves the price: AWS EU (Frankfurt) is $2.60 on Standard and $3.90 on Enterprise, and AWS Asia Pacific (Tokyo) is $2.85 and $4.30. Capacity customers apply the discount in their order form to these on-demand prices.
The standard warehouse table is the baseline. Snowflake also lists a Gen 2 warehouse table, where on AWS an X-Small uses 1.35 credits per hour and a Medium 5.4, so Gen 2 is not simply a cheaper standard warehouse; compare cost per finished query, not per hour.
BigQuery compute pricing: on-demand versus editions
The figures below are those the BigQuery pricing page shows first, for Iowa (us-central1). Google prices other regions differently, so pick your region on the page.
- On-demand$6.25 per TiB scanned; first 1 TiB per month free
- Standard edition (pay as you go)$0.04 per slot-hour
- Enterprise edition$0.06 pay as you go, $0.054 with a 1-year commitment, $0.048 with a 3-year commitment
- Enterprise Plus edition$0.10 pay as you go, $0.09 with a 1-year commitment, $0.08 with a 3-year commitment
Capacity is billed per second with a one-minute minimum by default, and you can opt in to fluid scaling for per-second billing with no minimum. Slot commitments are offered in Enterprise and Enterprise Plus, in a 50-slot minimum and increments of 50 slots. On-demand projects generally get access to up to 2,000 concurrent slots shared across all queries in the project.
Redshift compute pricing: provisioned versus Serverless
The Redshift pricing page says provisioned starts at $0.543 per hour and Serverless begins at $1.50 per hour. Serverless capacity is set as a base of 4 to 1024 RPUs, each with 16 GB of memory.
- Serverless rate. AWS prices it at $0.375 per RPU-hour in its US East (N. Virginia) examples. Concurrency scaling and Redshift Spectrum are included, with no separate charge.
- Provisioned rate. AWS example uses ra3.4xlarge nodes at $3.26 per hour, so four nodes over 730 hours is $9,519.20 a month before storage. Reserved Instances lower this; the page gives no percentage.
- Serverless Reservations. These reduce Serverless compute costs by up to 45%.
Storage pricing, briefly
Compute is not the whole bill, and the three storage meters differ:
- Snowflake: $23.00 per TB per month on-demand in AWS US East (N. Virginia), based on the average terabytes stored over the month.
- BigQuery: $0.000031507 per GiB-hour for active logical storage, which works out to about $0.023 per GiB-month by our arithmetic. Long-term logical storage is $0.000021918 per GiB-hour, and the first 10 GiB each month is free.
- Redshift: managed storage at $0.024 per GB-month in US East (N. Virginia), billed separately from compute on RA3 nodes.
A worked comparison from published rates
Assumptions: the compute is busy 8 hours a day for 22 days, or 176 hours a month; Snowflake is in AWS US East (N. Virginia); the BigQuery figures are the Iowa list prices; Redshift is in US East (N. Virginia). Nothing here is storage, and nothing assumes discounts or commitments.
- Snowflake Medium warehouse (4 credits per hour). Standard: 4 x $2.00 = $8.00 per hour, or $1,408 a month. Enterprise: 4 x $3.00 = $12.00 per hour, or $2,112.
- Redshift Serverless at 32 RPUs. 32 x $0.375 = $12.00 per hour, or $2,112 a month.
- BigQuery capacity at 100 slots, pay as you go. Enterprise: 100 x $0.06 = $6.00 per hour, or $1,056. Standard: 100 x $0.04 = $4.00 per hour, or $704.
- BigQuery on-demand, 50 TiB scanned in a month. (50 - 1) x $6.25 = $306.25, whatever the run time.
Cost-control features
- Snowflake. Warehouses auto-suspend when inactive and auto-resume on demand, both enabled by default (warehouse overview). Resource monitors trigger alerts or suspend warehouses at a percentage of a credit quota, and only ACCOUNTADMIN can create them. Multi-cluster warehouses are an Enterprise feature.
- BigQuery. The cost-control guide recommends dry runs and the query validator to estimate bytes, a maximum bytes billed setting per query, and custom daily quotas per project or user. It also notes that LIMIT will not cut compute costs on non-clustered tables.
- Redshift. Serverless has a Max setting in RPU-hours with a daily, weekly or monthly duration and actions to take when it is reached; you can also set the base capacity, or a price-performance target (Redshift pricing).
Which fits which workload
- Spiky, ad hoc analysis on large tables: BigQuery on-demand charges only for bytes read, so idle time is free and the first 1 TiB each month is free. Partitioning and clustering reduce what you scan.
- Steady, heavy query load: BigQuery capacity with commitments, or Redshift provisioned with Reserved Instances, give a fixed or discounted hourly rate.
- Many teams sharing one platform with separate budgets: Snowflake lets you size one warehouse per team and cap each with a resource monitor.
- AWS-centric stacks with variable load: Redshift Serverless bills only while queries run, with Serverless Reservations up to 45% off.
This is a pricing view. Choose with a proof of concept on your own data, because query speed and data layout change how many credits, slots or RPUs a job needs.
Questions people ask
Is Snowflake more expensive than BigQuery or Redshift?
It depends on the workload. On list price, a Snowflake Medium warehouse on Standard costs $8.00 per hour in AWS US East (N. Virginia), against $12.00 per hour for 32 RPUs of Redshift Serverless and $6.00 per hour for 100 BigQuery Enterprise slots. Those units are not equivalent, so test your own queries.
What does BigQuery on-demand cost per query?
It charges by data processed, not by query. The pricing page lists $6.25 per TiB after the first free 1 TiB each month, for the region it shows first.
How is a Snowflake credit priced?
By edition, cloud and region. In AWS US East (N. Virginia) it is $2.00 on Standard, $3.00 on Enterprise and $4.00 on Business Critical, per the Service Consumption Table.
Does Redshift Serverless charge when idle?
AWS says you pay in RPU-hours per second for the workloads you run, with a 60-second minimum and no charge for start-up time, per the pricing page.
How can I cap spend on each platform?
Snowflake uses resource monitors and auto-suspend, BigQuery uses maximum bytes billed and custom quotas, and Redshift Serverless uses a Max RPU-hours setting.
Do these prices include storage?
No. Storage is billed separately on all three: $23.00 per TB per month on Snowflake, $0.024 per GB-month on Redshift managed storage, and an hourly per-GiB rate on BigQuery.
Sources
- Snowflake Service Consumption Table (effective October 15, 2026) · credit prices by edition and region, warehouse sizes, storage
- Snowflake docs: Overview of warehouses · auto-suspend, multi-cluster
- Snowflake docs: Resource monitors · credit quotas and actions
- Google Cloud: BigQuery pricing · on-demand per TiB, editions per slot-hour, storage
- Google Cloud: Control BigQuery costs · dry runs, maximum bytes billed, custom quotas
- AWS: Amazon Redshift pricing · Serverless RPU-hours, provisioned nodes, managed storage
Prices and editions change often. Figures here were read from the linked vendor pages on October 17, 2026; confirm against the current pages and your own contract before you commit budget.
Top photo: Lydia Pintscher & Philippe Saadé / CC0 via Wikimedia Commons