StartStart using SigmaSigma templatesAI cost templates

Snowflake Cortex cost template

View as Markdown

Use the Snowflake Cortex cost template to monitor spend and usage across for all Snowflake Cortex features, including Cortex agents, AISQL, AI functions, Cortex Analyst, Cortex Code (CoCo), Cortex CoWork (Snowflake Intelligence), and Cortex Search.

In the template, you can review an executive cost overview, a per-product cost breakdown, and per-user and per-department adoption.

The template is set up with sample data, so you can explore the analysis before you connect your own data.

This template tracks usage across all Snowflake Cortex products. If you only want to track usage of Cortex AI Functions, see the Snowflake AI spend template. To analyze general Snowflake warehouse, cost, and performance data, see the Snowflake usage templates.

System and user requirements

The ability to use the Snowflake Cortex cost template requires the following:

  • You must be assigned an account type with the Create, edit, and publish workbooks permission enabled.
  • To use your own data, your organization must be connected to Snowflake and authenticated with a role that has privileges to view the relevant ACCOUNT_USAGE views. For details on the views, see Data sources.
  • To attribute spend and usage to people and departments, you must supply a users table that maps each Snowflake user to a person and department.

Data sources

The template uses sample data, but if you want to use your own data, the template visualizes your Snowflake account’s Cortex usage using the following ACCOUNT_USAGE views:

TableDetails
CORTEX_AGENT_USAGE_HISTORYCredits and tokens per agent invocation for Cortex Agents.
CORTEX_AISQL_USAGE_HISTORYCredits and tokens per AISQL function call (for example, AI_COMPLETE, AI_CLASSIFY, AI_FILTER), by model, function, and warehouse.1
CORTEX_AI_FUNCTIONS_USAGE_HISTORYCredits and tokens per AI function call (for example, COMPLETE, SUMMARIZE, EXTRACT), by model, function, and warehouse.
CORTEX_ANALYST_USAGE_HISTORYCredits and request volume for natural-language-to-SQL requests in Cortex Analyst.
Cortex Code usageCredits and token usage for Cortex Code usage from the CLI and Snowsight, including extensions that use the agent client protocol to access the CLI. Union the following tables together: CORTEX_CODE_CLI_USAGE_HISTORY (CLI), CORTEX_CODE_SNOWSIGHT_USAGE_HISTORY (Snowsight) and CORTEX_CODE_DESKTOP_USAGE_HISTORY (Desktop).
SNOWFLAKE_INTELLIGENCE_USAGE_HISTORYCredits and usage for Snowflake Intelligence (Cortex CoWork) interactions.
CORTEX_SEARCH_DAILY_USAGE_HISTORYDaily serving and indexing consumption for search services using Cortex Search.
METERING_DAILY_HISTORYAccount-level daily credits by service type, used to identify the total billed Cortex credits (CREDITS_USED and CREDITS_BILLED).2
Your users tableJoins to the per-product usage views on the Snowflake user identifier to attribute spend and usage to people and departments. Replaces the sample users table.3

1The AISQL and AI Functions views overlap. CORTEX_AI_FUNCTIONS_USAGE_HISTORY covers all AI function calls, including AI_PARSE_DOCUMENT, while CORTEX_AISQL_USAGE_HISTORY covers only the subset of AI functions used by AISQL. The template reports each product from its own view. To see a combined function total, use the METERING_DAILY_HISTORY total instead of adding the totals from the AISQL and AI Functions usage history views.

2Cortex spend includes several service types. The AI_SERVICES service type includes Cortex AI Functions and Analyst. Cortex Code bills separately under CORTEX_CODE_CLI and CORTEX_CODE_SNOWSIGHT. The template includes each Cortex service type in the total. Snowflake documents the service type list as partial, so if you add a Cortex product, confirm which service type it bills under before you rely on the top-line total.

3Snowflake does not provide person or department attributes. Some Cortex usage views identify each user only by a privacy-preserving hash. To attribute spend and usage to people and departments, replace the sample users table with your own users table that maps the Snowflake user identifier to a person and department. You can change the source to your table, or load your users into an input table with the same shape, keeping the column names identical so the joins and elements continue to work.

For further details about the data included in the Snowflake usage views, review the relevant Snowflake documentation pages.

Cortex bills in credits, not dollars. The workbook converts credits to USD by multiplying by a credit price — default $2.00 (global routing) or $2.20 (regional). Set the Credit price (USD / AI Credit) filter to your organization’s contracted rate and confirm your routing before treating the dollar figures as authoritative.

Create a workbook from the template

To create a workbook from this template:

  1. Go to Home > Templates, then select the Snowflake Cortex Cost template to review its details.

  2. Click Use template. The template opens as an exploration named after the template using sample data.

  3. To use your own data, click Change sources:

    In the Swap data sources overview modal, Sigma attempts to match each source to your own tables. Because the column names match your Cortex usage views, most sources match automatically.

    1. To replace a matched source, select the source drop-down menu, then select a new source.
    2. Match the sample users table to your own users table.
    3. Click Swap now, then explore the changes to confirm your data looks correct.

  4. In the Filters menu, set the Credit price (USD / AI Credit) to your contracted rate and the Date Range to the period you care about.

  5. To create an editable, publishable document, in the document header, click Save as, then name and save the workbook.

A Cortex product appears empty if your account has no usage for it in the selected period, or if that product’s usage view is not available in your Snowflake account. If you see data is missing, confirm the product is enabled in Snowflake and the corresponding ACCOUNT_USAGE view has data that you can access from Sigma.

For more details, see Create documents from templates.

About this template

The template contains Executive overview, Cost by product, and Adoption & users pages, and a hidden Data page for source data. The template also includes a README page to document the purpose of the template, source tables, and setup steps.

Executive overview

This page answers “How much are we spending on Snowflake Cortex, and where is it going?” and includes:

  • Total Cortex credits, Total Cortex spend, Active users, and Total calls KPIs, each with period-over-period change details.
  • A spend trend over time from the billed METERING_DAILY_HISTORY top-line total.
  • Spend by product details across the seven Cortex products (Agents, AISQL, AI Functions, Analyst, Code (CoCo), CoWork, and Search).
  • A snapshot of top departments and top users across all products.

Cost by product

This page answers “Which Cortex products and models drive spend?” and includes:

  • Credits and dollars by product, sorted by spend.
  • Per-product cost drivers, such as top models and functions for AISQL, AI Functions, and Agents, request volume for Analyst, and serving consumption for Search.
  • Cortex Code (CoCo) cost across its client surfaces (CLI and Snowsight).

Adoption & users

This page answers “Who is using Cortex?” and includes:

  • Active-user trend and adoption by department and by role.
  • A department-by-product matrix of usage across the Cortex products.
  • A top-users detail table of per-user credits, tokens, and calls.