Use an enterprise-grade service for the end-to-end machine learning lifecycle
Empower developers and data scientists with a wide range of productive experiences for building, training, and deploying machine learning models faster. Accelerate time to market and foster team collaboration with industry-leading MLOps—DevOps for machine learning. Innovate on a secure, trusted platform, designed for responsible AI.
Explore pricing options
-
Pay as you go
Pay for compute capacity by the second, with no long-term commitments or upfront payments. Increase or decrease consumption on demand.
-
Azure savings plan for compute
Save money across select compute services globally by committing to spend a fixed hourly amount for 1 or 3 years, unlocking lower prices until you reach your hourly commitment. Suited for dynamic workloads while accommodating for planned or unplanned changes.
-
Reservations
Azure Reserved Virtual Machine Instances provide significant cost reduction, compared to pay-as-you-go rates, when you commit to one-year or three-year terms. Suited for stable, predictable workloads with no planned changes.
Apply filters to customize pricing options to your needs.
Prices are estimates only and are not intended as actual price quotes. Actual pricing may vary depending on the type of agreement entered with Microsoft, date of purchase, and the currency exchange rate. Prices are calculated based on US dollars and converted using London closing spot rates that are captured in the two business days prior to the last business day of the previous month end. If the two business days prior to the end of the month fall on a bank holiday in major markets, the rate setting day is generally the day immediately preceding the two business days. This rate applies to all transactions during the upcoming month. Sign in to the Azure pricing calculator to see pricing based on your current program/offer with Microsoft. Contact an Azure sales specialist for more information on pricing or to request a price quote. See frequently asked questions about Azure pricing.
Please note, there is no additional charge to use Azure Machine Learning. However, along with compute, you will incur separate charges for other Azure services consumed, including but not limited to Azure Blob Storage, Azure Key Vault, Azure Container Registry and Azure Application Insights.
To learn more about how to manage budgets, costs, and quota for Azure Machine Learning, see here.
General purpose
Balanced CPU-to-memory ratio. Ideal for testing and development, small to medium databases, and low to medium traffic web servers.
D2-64 v3
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| D2 v3 | 2 | 8 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D4 v3 | 4 | 16 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D8 v3 | 8 | 32 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D16 v3 | 16 | 64 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D32 v3 | 32 | 128 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D64 v3 | 64 | 256 GiB | $- |
$-
|
$-
|
$-
|
$-
|
D2s-64s v3
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| D2s v3 | 2 | 8 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D4s v3 | 4 | 16 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D8s v3 | 8 | 32 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D16s v3 | 16 | 64 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D32s v3 | 32 | 128 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D64s v3 | 64 | 256 GiB | $- |
$-
|
$-
|
$-
|
$-
|
D1-5 v2
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| D1 v2 | 1 | 3.5 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D2 v2 | 2 | 7 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D3 v2 | 4 | 14 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D4 v2 | 8 | 28 GiB | $- |
$-
|
$-
|
$-
|
$-
|
D1s-5s v2
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| DS1 v2 | 1 | 3.5 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| DS2 v2 | 2 | 7 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| DS3 v2 | 4 | 14 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| DS4 v2 | 8 | 28 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| DS5 v2 | 16 | 56 GiB | $- |
$-
|
$-
|
$-
|
$-
|
D2ds – D64ds v4
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| D2ds v4 | 2 | 8 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D4ds v4 | 4 | 16 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D8ds v4 | 8 | 32 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D16ds v4 | 16 | 64 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D32ds v4 | 32 | 128 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D48ds v4 | 48 | 192 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D64ds v4 | 64 | 256 GiB | $- |
$-
|
$-
|
$-
|
$-
|
Compute optimized
High CPU-to-memory ratio. Good for medium traffic web servers, network appliances, batch processes, and application servers.
Fsv2-series
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| F2s v2 | 2 | 4 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| F4s v2 | 4 | 8 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| F8s v2 | 8 | 16 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| F16s v2 | 16 | 32 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| F32s v2 | 32 | 64 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| F64s v2 | 64 | 128 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| F72s v2 | 72 | 144 GiB | $- |
$-
|
$-
|
$-
|
$-
|
Memory optimized
High memory-to-core ratio. Great for relational database servers, medium to large caches, and in-memory analytics.
E2-64 v3
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| E2 v3 | 2 | 16 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E4 v3 | 4 | 32 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E8 v3 | 8 | 64 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E16 v3 | 16 | 128 GiB | $- |
$-
|
$-
|
$-
|
$-
|
D-series
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| D1 | 1 | 3.5 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D2 | 2 | 7 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D3 | 4 | 14 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D4 | 8 | 28 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D11 | 2 | 14 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D12 | 4 | 28 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D13 | 8 | 56 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D14 | 16 | 112 GiB | $- |
$-
|
$-
|
$-
|
$-
|
D11-15 v2
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| D11 v2 | 2 | 14 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D12 v2 | 4 | 28 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D13 v2 | 8 | 56 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D14 v2 | 16 | 112 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| D15 v2 | 20 | 140 GiB | $- |
$-
|
$-
|
$-
|
$-
|
D11S-15S v2
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| DS11 v2 | 2 | 14 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| DS12 v2 | 4 | 28 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| DS13 v2 | 8 | 56 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| DS14 v2 | 16 | 112 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| DS15 v2 | 20 | 140 GiB | $- |
$-
|
$-
|
$-
|
$-
|
E2a – E96a v4
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| E2a v4 | 2 | 16 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E4a v4 | 4 | 32 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E8a v4 | 8 | 64 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E16a v4 | 16 | 128 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E32a v4 | 32 | 256 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E48a v4 | 48 | 384 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E64a v4 | 64 | 512 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| E96a v4 | 96 | 672 GiB | $- |
$-
|
$-
|
$-
|
$-
|
E2s – E64s
Spark compute from Synapse. Pricing is based on Synapse pricing. For more details, please refer to Pricing - Azure Synapse Analytics | Microsoft Azure (Big Data Analytics section).
M-series
| Instance | vCPU(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| M64 | 64 | 1,024 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M32ls | 32 | 256 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M64ls | 64 | 512 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M64m | 64 | 1,792 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M128m | 128 | 3,892 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M8ms | 8 | 218.75 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M16ms | 16 | 437.5 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M32ms | 32 | 875 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M64ms | 64 | 1,792 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M128ms | 128 | 3,892 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M64s | 64 | 1,024 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M128 | 128 | 2,048 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M128s | 128 | 2,048 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| M32ts | 32 | 192 GiB | $- |
$-
|
$-
|
$-
|
$-
|
GPU
Specialized virtual machines targeted for heavy graphic rendering and video editing available with single or multiple GPUs.
NC-series
| Instance | vCPU(s) | RAM | GPU |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|---|
| NC6 | 6 | 56 GiB | 1X K80 | $- |
$-
|
$-
|
$-
|
$-
|
| NC12 | 12 | 112 GiB | 2X K80 | $- |
$-
|
$-
|
$-
|
$-
|
| NC24 | 24 | 224 GiB | 4X K80 | $- |
$-
|
$-
|
$-
|
$-
|
| NC24r | 24 | 224 GiB | 4X K80 | $- |
$-
|
$-
|
$-
|
$-
|
NCsv2-series
| Instance | vCPU(s) | RAM | GPU |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|---|
| NC6s v2 | 6 | 112 GiB | 1X P100 | $- |
$-
|
$-
|
$-
|
$-
|
| NC12s v2 | 12 | 224 GiB | 2X P100 | $- |
$-
|
$-
|
$-
|
$-
|
| NC24s v2 | 24 | 448 GiB | 4X P100 | $- |
$-
|
$-
|
$-
|
$-
|
| NC24rs v2 | 24 | 448 GiB | 4X P100 | $- |
$-
|
$-
|
$-
|
$-
|
NCsv3-series
| Instance | vCPU(s) | RAM | GPU |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|---|
| NC6s v3 | 6 | 112 GiB | 1X V100 | $- |
$-
|
$-
|
$-
|
$-
|
| NC12s v3 | 12 | 224 GiB | 2X V100 | $- |
$-
|
$-
|
$-
|
$-
|
| NC24s v3 | 24 | 448 GiB | 4X V100 | $- |
$-
|
$-
|
$-
|
$-
|
| NC24rs v3 | 24 | 448 GiB | 4X V100 | $- |
$-
|
$-
|
$-
|
$-
|
NV-series
| Instance | vCPU(s) | RAM | GPU |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|---|
| NV6 | 6 | 56 GiB | 1X M60 | $- |
$-
|
$-
|
$-
|
$-
|
| NV12 | 12 | 112 GiB | 2X M60 | $- |
$-
|
$-
|
$-
|
$-
|
| NV24 | 24 | 224 GiB | 4X M60 | $- |
$-
|
$-
|
$-
|
$-
|
NVv3-series
| Instance | vCPU(s) | RAM | GPU |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|---|
| NV12s v3 | 12 | 112 GiB | 1X M60 | $- |
$-
|
$-
|
$-
|
$-
|
| NV24s v3 | 24 | 224 GiB | 2X M60 | $- |
$-
|
$-
|
$-
|
$-
|
| NV48s v3 | 48 | 448 GiB | 4X M60 | $- |
$-
|
$-
|
$-
|
$-
|
NDs-series
| Instance | Core(s) | RAM | GPU |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|---|
| ND6s | 6 | 112 GiB | 1X P40 | $- |
$-
|
$-
|
$-
|
$-
|
| ND12s | 12 | 224 GiB | 2X P40 | $- |
$-
|
$-
|
$-
|
$-
|
| ND24rs | 24 | 448 GiB | 4X P40 | $- |
$-
|
$-
|
$-
|
$-
|
| ND24s | 24 | 448 GiB | 4X P40 | $- |
$-
|
$-
|
$-
|
$-
|
NCas_T4_v3-series
| Instance | vCPU(s) | RAM | GPU |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|---|
| NC4as T4 v3 | 4 | 28 GiB | 1X T4 | $- |
$-
|
$-
|
$-
|
$-
|
| NC8as T4 v3 | 8 | 56 GiB | 1X T4 | $- |
$-
|
$-
|
$-
|
$-
|
| NC16as T4 v3 | 16 | 110 GiB | 1X T4 | $- |
$-
|
$-
|
$-
|
$-
|
| NC64as T4 v3 | 64 | 440 GiB | 4X T4 | $- |
$-
|
$-
|
$-
|
$-
|
NDv2-series
| Instance | Core(s) | RAM | GPU |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|---|
| ND40rs v2 | 40 | 672 GiB | 8X V100 (NVlink) | $- |
$-
|
$-
|
$-
|
$-
|
ND A100 v4 series
| Instance | vCPU(s) | RAM | GPU |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|---|
| ND96asr A100 v4 | 96 | 900 GiB | 8x A100 (NVlink) | $- |
$-
|
$-
|
$-
|
$-
|
High performance compute
Our fastest and most powerful CPU virtual machines with optional high-throughput network interfaces (RDMA).
H-series
| Instance | Core(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| H8 | 8 | 56 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| H8m | 8 | 112 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| H16m | 16 | 224 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| H16mr | 16 | 224 GiB | $- |
$-
|
$-
|
$-
|
$-
|
| H16r | 16 | 112 GiB | $- |
$-
|
$-
|
$-
|
$-
|
HB-series
| Instance | Core(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| HB60rs | 60 | 228 GiB | $- |
$-
|
$-
|
$-
|
$-
|
HBv2-series
| Instance | Core(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| HB120rs v2 | 120 | 456 GiB | $- |
$-
|
$-
|
$-
|
$-
|
HC-series
| Instance | Core(s) | RAM |
Linux VM Price
|
Machine Learning Service Surcharge |
Pay As You Go Total Price |
1 year savings plan | 3 year savings plan |
|---|---|---|---|---|---|---|---|
| HC44rs | 44 | 352 GiB | $- |
$-
|
$-
|
$-
|
$-
|
Managed Spark
(Instance types: E2s-E64s)
Managed Spark from Azure ML follows the same pricing found in Synapse Big Data Analytics (Spark pools):
| Type | Price |
|---|---|
| Memory Optimized | $- per vCore-hour |
Azure pricing and purchasing options
Connect with us directly
Get a walkthrough of Azure pricing. Understand pricing for your cloud solution, learn about cost optimization and request a custom proposal.
Talk to a sales specialistSee ways to purchase
Purchase Azure services through the Azure website, a Microsoft representative, or an Azure partner.
Explore your optionsAdditional resources
Azure Machine Learning
Learn more about Azure Machine Learning features and capabilities.
Pricing calculator
Estimate your expected monthly costs for using any combination of Azure products.
SLA
Review the Service Level Agreement for Azure Machine Learning.
Documentation
Review technical tutorials, videos, and more Azure Machine Learning resources.
Frequently asked questions
-
Azure Machine Learning customers are responsible for the costs of Azure resources consumed including Virtual Machines.
-
You will be billed daily. For billing purposes, a day commences at midnight UTC. Bills are generated monthly.
-
Training:
As a specific example, let’s say you train a model for 100 hours using 10 DS14 v2 VMs on an Basic workspace in US West 2. For a billing month of 30 days, your bill will be as follows:
Azure VM Charge: (10 machines * $1.196 per machine) * 100 hours = $1196
Azure Machine Learning Charge: (10 machines * 16 cores * $0 per core) * 100 hours = $0
Total: $1196 + $0 = $1196
-
Inferencing:
As a specific example, let’s say you deploy a model for inferencing all day for a 30-day billing month using 10 DS14 v2 VMs in Basic in US West 2. For a billing month of 30 days, your bill will be as follows:
Azure VM Charge: (10 machines * $1.196 per machine) * (24 hours * 30 days) = $8611.2
Azure Machine Learning Charge: (10 machines * 16 cores * $0 per core) * (24 hours * 30 days) = $0
Total: $8611.2 + $0 = $8611.2
Please note there are no additional Azure Machine Learning charges. Along with compute charges, you will incur separate charges for any Azure services consumed, including but not limited to HDInsight, Azure Container Registry, Azure Blob Storage, Application Insights, Azure Key Vault, Virtual Network, Azure Event Hub, and Azure Stream Analytics.
-
Training:
Talk to a sales specialist for a walk-through of Azure pricing. Understand pricing for your cloud solution.
Get free cloud services and a $200 credit to explore Azure for 30 days.