Databricks Cost Calculator

See the full cost of a Databricks workload, including the cloud VMs underneath, at list or your own DBU rate.

Multi-cloudCalculator60 seconds

New here? See what this tool does and how to use it.

Start from a scenario

Each one fills in the fields below for a typical workload of that kind. You can change any of them.

Cloud

Plan

Compute type

Interactive notebooks. The expensive one. List rate $0.55 per DBU (AWS Premium, US East (N. Virginia)).

Your discount off listFrom your contract. 0 for list price.

Your rate per DBU (optional)Replaces list less discount for this compute type. 0 uses list. The Jobs comparison takes the same share of list.

USD

Instance typeSets DBUs per hour from Databricks' list and the EC2 on-demand price in us-east-1.

Worker nodesA driver node is added automatically and bills the same.

Cluster hours per month

Idle hours a dayWithin the hours above: time it runs with nobody using it, such as overnight. Counted over 30.44 days.

Workers on spotThe driver always stays on-demand.

Monthly Databricks cost$2,057.40

DBUs: $1,366.20
Cloud instances underneath: $691.20
True cost per DBU dollar: 1.51x

$24,688.80 a year at these inputs.

Estimate coverage: medium. DBU rates are Databricks’ list rates for the cloud and plan picked (US East (N. Virginia)), less your discount. On AWS the DBUs per hour and the EC2 on-demand price come from the instance type’s row. The spot discount and the Photon speed-up are planning assumptions: check them against your own bills.

True monthly cost

$2,057.40

Your Databricks invoice will show $1,366.20. The other $691.20 arrives from your cloud provider, a total of 1.5x the Databricks line.

Databricks DBUs2,484 DBUs at $0.55, 66.4% of the total shown	$1,366.20
Cloud instances underneath9 nodes (8 workers + 1 driver) x 400 hours at $0.19 a node-hour	$691.20
Total	$2,057.40

Move this workload to Jobs Compute

Monthly saving from the compute type alone

$993.60

Identical work, identical cluster, $1,063.80 instead of $2,057.40. All-Purpose exists for interactive notebooks; anything on a schedule belongs on Jobs.

Other levers

Spot workers

$0.00/mo

At a 70% spot discount. Applies to the cloud half only; DBUs cost the same on spot.

DBU share of the bill

66.4%

If you budgeted from the Databricks invoice alone, this is the fraction you saw.

Effective DBU rate

$0.55

The All-Purpose Compute list rate on AWS Premium.

What this model assumes

Show

The cloud VMs are the point. Databricks bills DBUs; your cloud provider separately bills the instances those DBUs ran on. A calculator that prices only DBUs leaves that second invoice out. How large it is depends on the compute type, the instance and your DBU rate, so the split for your inputs is shown above rather than assumed.
List rates per compute type, less your discount. Each compute type has its own line on Databricks’ rate card, and the lines move differently by cloud and plan: Jobs Compute is $0.15 on AWS Premium and $0.30 on Azure Premium, while All-Purpose is $0.55 on both.
The driver node is counted. It runs and bills like a worker, and it cannot go on spot.
Photon is modeled as faster, not just pricier. Databricks lists Photon at 2.9x the DBUs an hour on Jobs Compute and 2x on All-Purpose, and it finishes sooner, by 2.0x here (2x unless you change it), so the net can go either way. The speed-up is an assumption: measure yours before trusting the direction.
Spot workers are discounted by 70%. An assumption (70% unless you change it): real spot prices move daily and differ by instance type and zone.
SQL warehouses are priced by size. Each size has a fixed DBU rate an hour (Small 12, Medium 24). Serverless includes the compute. Classic and Pro also run VMs in your cloud account, which this page does not price.
Idle hours and auto-termination are simplified. The idle hours are one stretch a day, and auto-termination bills that stretch only until its timer runs out.
Not covered: Jobs Serverless, serverless notebooks, DLT and Lakeflow pipelines, model serving and AI functions, storage, networking and Unity Catalog.

Rates used and their sourcesShow

Databricks list DBU rates by cloud, plan and compute type (rate card), source, read 23 Oct 2026
Databricks DBUs per hour by AWS instance type, source, read 23 Oct 2026
EC2 on-demand (Linux, shared tenancy): m5.xlarge and m5.2xlarge, source, read 13 Oct 2026
EC2 on-demand, us-east-1 Linux: m5.4xlarge, m5d.xlarge, m6gd.xlarge, c5.2xlarge, r5.xlarge, r5.2xlarge, r5d.xlarge, i3.xlarge, i3.2xlarge, i3.4xlarge, source, read 23 Oct 2026

Prices verified 13 Oct 2026. Public list prices in USD for US East (N. Virginia), before any discount you have negotiated.

Want this done on your real bill?

BigQuery Cost Audit: We find the queries, tables and slots driving your BigQuery bill and fix them without slowing anyone down. From $2,500, fixed. Nothing paid upfront.

See the service

About this tool

The DBU rate starts at Databricks' list rate for the cloud, plan and compute type you pick, and on AWS the DBUs per hour and the EC2 price come from the instance type. Enter your discount or your own rate: most contracts are not list.

Questions

What is a DBU and what does it actually cost?

A Databricks Unit is a normalized measure of processing consumed per hour. Databricks bills DBUs; your cloud provider separately bills the virtual machines those DBUs ran on, on an invoice the Databricks bill never shows. Budgeting from the Databricks bill alone leaves that second invoice out, and how large it is depends on the compute type, the instance size and your DBU rate. This calculator shows the split for your own inputs. Serverless compute is the exception: its DBU rate includes the compute, so there is no VM line.

Why is All-Purpose compute so much more expensive than Jobs compute?

Because it is priced for interactive use, a cluster you attach a notebook to and leave running. The DBU rate is several times the Jobs rate for identical work. Any workload on a schedule belongs on Jobs compute, and moving it is a configuration change rather than a rewrite. Interactive clusters left running overnight are the classic Databricks bill shock; notebooks cannot move to Jobs, so the fix there is auto-termination, which this page prices.

Does Photon save money on Databricks?

It can go either way. Databricks lists Photon at 2.9 times the DBUs an hour on Jobs Compute and twice on All-Purpose, at the same price per DBU, but it finishes the same work faster, so the net depends entirely on the speed-up your workload actually gets. For scan-heavy SQL it is frequently a win; for workloads dominated by Python UDFs it often is not. Measure the speed-up before assuming the direction.

Do spot instances reduce Databricks costs?

They reduce the cloud half only. DBU rates are the same whether the node is spot or on-demand, so the saving is the spot discount times the worker share of the VM bill, which this calculator shows for your inputs. The driver node cannot go on spot without risking the whole cluster, so it stays on-demand. Serverless compute has no VMs of yours to put on spot.

What DBU rate does this calculator use?

Databricks' own list rate for the cloud, plan and compute type you pick, for example $0.15 a DBU for Jobs Compute and $0.55 for All-Purpose on AWS Premium, or $0.30 and $0.55 on Azure Premium. Most contracts carry a discount: enter yours, or type your own rate per DBU.

What does this calculator not cover?

Jobs Serverless, serverless notebooks, DLT and Lakeflow pipelines, model serving and AI functions, storage and networking. SQL Classic and Pro warehouses are priced in DBUs only: their VMs run in your cloud account and are on that bill.

More tools for: Analytics and warehouses, Compute and utilization

Want this measured continuously, against real usage?

Finitizer does this for BigQuery today. Snowflake and Redshift support is on the way.

See how Finitizer works

Databricks Cost Calculator