Why is your Snowflake bill so high?

Paste one row per warehouse — size, credits used, AUTO_SUSPEND, resumes — and see what each warehouse costs, the idle tail an AUTO_SUSPEND of 600 seconds is billing, idle warehouses, warehouses that never suspend and the ready-to-paste ALTER WAREHOUSE command.

Nothing is uploaded and Snowflake is never contacted. The text box is analysed in this browser and never leaves your device. The AUTO_SUSPEND saving is modelled from resumes and idle tails, which is an estimate.

1. Paste your warehouses

Rows of warehouse, size, credits this month, AUTO_SUSPEND seconds (or never), queries, avg query seconds, resumes (last three optional)

Prefer a CLI, or want a CI check?

Same logic, zero dependencies, one command — reads a file or stdin:

node cli.mjs analyze warehouses.csv
cat warehouses.csv | node cli.mjs analyze -
node cli.mjs analyze warehouses.csv --suspend 120 --fail-on warn   # exit 1 on any waste
node cli.mjs analyze warehouses.csv --json --prices my-prices.json

Get cli.mjs and engine.mjs — Node 18+, nothing to install, no network call. More: Snowflake warehouse cost: AUTO_SUSPEND, idle warehouses and sizing.

What's recognised

Input	Handling

Prices are illustrative and dated; edit them with --prices in the CLI. Storage, serverless features, cloud-services credits and multi-cluster warehouses are not modelled.

Common questions

Does this connect to my Snowflake account?

No. It reads the rows you paste and never contacts Snowflake. It cannot see your contract, so every dollar figure is the credits you pasted times a credit price you can edit.

How do I get the numbers?

Credits per warehouse: SUM(CREDITS_USED) grouped by WAREHOUSE_NAME over 30 days from SNOWFLAKE.ACCOUNT_USAGE.WAREHOUSE_METERING_HISTORY. Size and AUTO_SUSPEND: SHOW WAREHOUSES. Queries and average runtime: ACCOUNT_USAGE.QUERY_HISTORY. Resumes: RESUME events in WAREHOUSE_EVENTS_HISTORY.

What AUTO_SUSPEND should I use?

Snowflake bills a 60-second minimum each time a warehouse resumes, so 60 seconds is the practical floor. Anything longer pays for idle time after each burst; the default of 600 seconds on a warehouse that wakes hundreds of times a month is a common silent cost. A shorter setting means a colder cache and a short resume wait.

Will a smaller warehouse be cheaper?

Credits per hour halve with each size down, but a long or memory-hungry query can run more than twice as long and cost the same or more. Short queries that do not speed up on a bigger size are the safe candidates. The advisor reports this only as an upper bound and never counts it in the saving.

Is my data uploaded anywhere?

No. The browser page analyses the text in your own browser and never sends it anywhere. The CLI reads a local file or stdin and makes no network call.