Warehouse Usage

The Warehouse usage page gives you a clear view of your Snowflake consumption without having to write SQL against Snowflake's account usage views. Use it to:

Track how many credits you've used (and been billed for) over any time period
See which warehouses, users and service categories are using the most credits
Find the specific queries that cost the most, so you know where to optimise
Spot when your warehouses are busiest across the week

Summary

Shows how many credits you use and what drives the spend.

Total Billed Credits: closest to your Snowflake invoice; includes compute, cloud services after discounts, and additional services such as AI and backup.
Total Credits: compute plus cloud-services adjustment; excludes additional services.
Compute Credits: warehouse usage for queries and workloads.
Cloud Services Credits: background services, charged only above the applicable threshold.
Credits Used: trends over time by warehouse.
Credits by Service Category: shows which Snowflake services drive spend; discounts may appear as negative values.
Credits by User and Warehouse: highlights the biggest consumers.

Account Usage

Identify costly queries and opportunities to reduce warehouse spend.

Query metrics: total queries, average runtime, bytes scanned and remote storage spill. Spill should ideally be 0 B; optimise affected queries or consider a larger warehouse.
Credits Breakdown: shows active, compilation, queue and idle usage. High idle usage suggests reviewing auto-suspend; high queue overload suggests capacity pressure.
Execution Time: shows daily query runtime by warehouse.
Query filters: narrow the query summaries, charts and cost table by query type, warehouse, user and Top N query shapes (similar query patterns grouped together).
Frequency vs Cost: prioritise queries that run often and consume the most credits. Optimise expensive runs or reduce unnecessary frequency.
Credits by Query Type: shows which statement types drive spend.
Higher Cost Queries: ranked by total credits burned. Start with the highest-cost shapes; compare credits per execution and bytes spilled to identify fixes.