Serverless DBU consumption by SKU

This article explains the SKUs and DBU multipliers used to bill for various Databricks serverless offerings.

For Azure Databricks pricing, see pricing details.

What is a DBU multiplier?

When using certain features, a multiplier is applied to the underlying DBUs consumed. For instance, Data Quality Monitoring has a 2X multiplier. If the associated background job uses 5 DBUs, you are billed for 10 DBUs after applying the multiplier. The DBUs shown on your bill and in system tables reflect the final amount after this multiplier is applied. See What is a DBU for the definition of a DBU.

Automated Serverless SKU

The following capabilities are billed against the Automated Serverless SKU.

Feature	DBU multiplier
Serverless Jobs	1X
Serverless Spark Declarative Pipelines	1X
Predictive Optimization	1X
Data Quality Monitoring	2X
Fine-Grained Access Control	1X
Online tables synchronization (Preview)	1X
Online tables Capacity Unit (Preview)	2X
Materialized Views and Streaming Tables in Databricks SQL	1X
Data Classification	3X

Interactive Serverless SKU

The following capabilities are billed against the Interactive Serverless SKU.

Product / Feature	DBU Multiplier
Serverless Notebook	1X
Databricks Apps - Medium compute/hr	0.5X
Databricks Apps - Large compute/hr	1X

Database Serverless Compute SKU

The following capabilities are billed against the Database Serverless SKU.

Product / Feature	DBU Multiplier
Lakebase Provisioned Compute	1X

SQL Serverless SKU

The following capabilities are billed against the SQL Serverless SKU.

Warehouse Size	DBU/hour
2X-Small	4
X-Small	6
Small	12
Medium	24
Large	40
X-Large	80
2X-Large	144
3X-Large	272
4X-Large	528

Model Serving SKU

The following capabilities are billed against the Serverless Real-Time Inference SKU.

AI Gateway

Product / Feature	DBU Multiplier
Inference Tables for CPU, GPU endpoints	7.143 DBUs / 1 GB of payload
Usage Tracking for CPU, GPU endpoints	1.429 DBUs / 1 GB of payload

CPU Model Serving

1 concurrent request/hr = 1 DBU/hr

GPU Model Serving

Instance Size	GPU configuration	DBUs / hour
Small	T4 or equivalent	10.48
XLarge	A100 80GB x 1 GPU or equivalent	78.60
2XLarge	A100 80GB x 2 GPU or equivalent	157.20
4XLarge	A100 80GB x 4 GPU or equivalent	314.40

Foundation Model Serving

Model	Pay-Per-Token: DBU / 1M INPUT tokens	Pay-Per-Token: DBU / 1M OUTPUT tokens	Provisioned Throughput: DBU per hour
Llama 4 Maverick	7.143	21.429	85.714
Llama 3.3 70B	7.143	21.429	85.714
GPT OSS 120B	2.143	8.571	71.429
Gemma 3 12B	2.143	7.143	71.429
Llama 3.1 8B	2.143	6.429	53.571
GPT OSS 20B	1.000	4.286	53.571
Llama 3.2 3B	n/a	n/a	46.429
Llama 3.2 1B	n/a	n/a	42.857
GTE	1.857	n/a	20.000
BGE Large	1.429	n/a	24.000

Anthropic Model Serving

Model	Endpoint Type	Context Length	Pay-Per-Token: DBU / 1M INPUT tokens	Pay-Per-Token: DBU / 1M OUTPUT tokens	Pay-Per-Token: DBU / 1M CACHE WRITES tokens	Pay-Per-Token: DBU / 1M CACHE READS tokens	Batch Inference: DBU per hour
Claude Opus 4.5	Global	Short Context	71.429	357.143	89.286	7.143	n/a
	In-geo		78.571	392.857	98.214	7.857	n/a
Claude Opus 4 / 4.1	Global/In-geo	All Lengths	214.286	1,071.43	267.857	21.429	514.286
Claude Sonnet 4.5	Global	Short Context	42.857	214.286	53.571	4.286	214.286
	In-geo		47.143	235.715	58.928	4.715	235.715
	Global	Long Context (>200k tokens)	85.714	321.429	107.143	8.571	214.286
	In-geo		94.285	353.572	117.857	9.428	235.715
Claude Sonnet 3.7 / 4 / 4.1	Global/In-geo	Short Context	42.857	214.286	53.571	4.286	214.286
		Long Context (>200k tokens)	85.714	321.429	107.143	8.571	214.286
Claude Haiku 4.5	Global	All Lengths	14.286	71.429	17.857	1.429	n/a
	In-geo		15.715	78.572	19.643	1.572	n/a

Shutterstock Image AI

1 image = 0.857 DBUs

Vector Search

Endpoint option	DBU/hour for 1 unit	Vector Capacity per Unit
Standard	4.0	2 million
Storage optimized	18.29	64 million

Model	DBUs per 1k requests
Databricks Vector Search Reranker	28.571

Agent Evaluation

Product	DBUs
Agent Evaluation LLM Judge	2.14 DBUs/M input tokens 8.57 DBUS/M output tokens
Agent Evaluation Synthetic Data	5.0 DBU per question generated

AI Functions

Server	Workload type	Estimated SRTI DBUs*
AI Parse Document	Simple page with no caption Simple page with captions Medium complexity page with tables, images, captions High complexity page with detailed diagrams and captions	10-15 DBUs 20-25 DBUs 60-65 DBUs 85-90 DBUs

* Estimated prices shown before 50% promotional discount running until June 30, 2026

Model Training

The following capabilities are billed against the Model Training SKU.

Model Training - Fine Tuning

Model	Training word count	Approximate DBUs	Approximate cost/run ($0.65/DBU US East)
Llama 3.3 70B	10,000,000	225	$146.25
	500,000,000	11,000	$7,150.00
Llama 3.1 70B	10,000,000	225	$146.25
	500,000,000	11,000	$7,150.00
Llama 3.1 8B	10,000,000	100	$65.00
	500,000,000	4,400	$2,860.00
Llama 3.2 3B	10,000,000	75	$48.75
	500,000,000	2,750	$1,787.50
Llama 3.2 1B	10,000,000	25	$16.25
	500,000,000	1,100	$715.00

Model Training - Forecasting

Feature	DBU Multiplier
Model Forecasting	4.0X

Databricks Storage

The following capabilities are billed against the Databricks Storage SKU

Product / Feature	DSU Multiplier
Vector Search	10X
Databricks Storage - Per GB of stored data	1X
Databricks Storage - Per 1000 write operations	0.3535X
Databricks Storage - Per 1000 read operations	0.0226X

Feedback

Was this page helpful?

Last updated on 2025-12-19