Creating a Metric

There is no single "create a metric" recipe — which endpoints to call first depends on where the data lives. Pick a row below, then open the linked guide.

You have...	Workflow
A metric already defined in a dbt semantic layer	Importing a dbt metric
An existing metric (or metrics) you want to combine/transform	Creating a calculated metric
Data in a data warehouse or a Cube deployment	Creating a metric from a data warehouse or Cube connection
Data in a supported SaaS service (Salesforce, GA4, ...)	Creating a metric from a service
Data behind a REST URL you can point at	Creating a metric from a REST endpoint
Raw data in hand (a CSV, a spreadsheet, values you already computed)	Creating a metric from uploaded data

All six end at POST /metrics; what differs is what you call first to arrive at a valid metric spec. See Getting Started for auth, base URL, and response conventions, and the Metrics guide for the full field reference on the resource itself.