Here’s a breakdown of Azure Databricks pricing, along with factors that can influence your costs:

Pricing Models

Azure Databricks offers flexible pricing models to fit your needs:

• Pay-as-You-Go: The most flexible model. You’re charged for:
  • Virtual Machines (VMs) running your clusters.
  • Databricks Units (DBUs): Units of processing capacity are billed per second based on the instance type you choose.
• Commit Plans:  Offers discounted rates (up to 37%) in exchange for committing to a specific DBU usage for a 1- or 3-year term.

Cost Factors

• VM Instance Types: The power of the instance types (Standard, Premium, etc.) will influence the cost. More powerful instances cost more per hour.
  • Compute Types: All-Purpose Compute: Best for general interactive workloads, data science, and machine learning.
  • Jobs Compute: Cost-effective for automated production ETL jobs.
  • Jobs Light Compute: Best for smaller, less demanding ETL jobs.
• Auto-termination: Turn on auto-termination to save costs when clusters are idle.
• Region: Azure Databricks pricing can differ slightly based on the area.

Pricing Examples (approximate, can vary based on region)

Compute TypeStandard Instance (DBU/hr)Premium Instance (DBU/hr)

All-Purpose Compute

$0.40

$0.55

Jobs Compute

$0.15

$0.30

Jobs Light Compute

$0.07

$0.22

drive_spreadsheetExport to Sheets

How to Optimize Your Costs

• Choose the suitable instance types:  Select more minor instances for less demanding tasks.
• Use auto-scaling:  Automatically adjust cluster size based on workload.
• Turn on auto-termination: Avoid paying for idle clusters.
• Consider Commit Plans: Commit Plans offer savings if you have predictable usage.
• Spot Instances: Consider using Azure Spot Instances for certain workloads to save costs.