Tool profile
Azure Data Factory: cost, fit and limits
Azure Data Factory is a managed data integration and orchestration service from Microsoft Azure that builds pipelines and code-free data flows across more than 90 connectors, runs SSIS packages and bills orchestration per 1,000 activity runs, trigger executions and debug runs.
What to know first
Azure Data Factory charges per 1,000 runs for orchestration across three integration runtimes: Azure, Managed VNET and Self-Hosted. The pricing page we opened showed placeholders instead of numeric rates, so no price is quoted here. The first five low-frequency activities a month are free, and usage past 100 activities gets a 20 percent discount. The product page names no AI feature.
Azure Data Factory at a glance
| Maker | Microsoft (Azure), Not stated on the vendor pages |
|---|---|
| Free plan | Yes |
| Entry price | Quote only |
| API access | No |
| Category | Best AI Data Orchestration Tools in 2026, rank 10 of 12 |
| Website | azure.microsoft.com/en-us/products/data-factory |
Who Azure Data Factory is best for
Microsoft-centered data teams
Connectors reach Azure Synapse Analytics and SQL Server beside Amazon Redshift, Google BigQuery, Salesforce, ServiceNow, Oracle Exadata and Teradata, so one service can bridge several vendors' systems.
Teams migrating SSIS packages
The product page says SSIS can be rehosted and extended, with up to 88 percent cost savings claimed when Azure Hybrid Benefit applies.
Analysts who prefer visual tools
Code-free data flows with intent-driven mapping let people build transformations without writing code, then run them inside pipelines, which keeps simple transformations out of hand-written code.
Azure Data Factory plans and pricing
| Plan | Price | What it includes |
|---|---|---|
| Azure Integration Runtime | Per 1,000 runs, rate not readable on the captured page | Cloud orchestration covering activity runs, trigger executions and debug runs, with the first 5 low-frequency activities each month free. |
| Managed VNET Integration Runtime | Per 1,000 runs, rate not readable on the captured page | Pipelines inside a managed virtual network, with up to 50 concurrent activities listed on the pricing page. |
| Self-Hosted Integration Runtime | Per 1,000 runs, rate not readable on the captured page | On-premises orchestration with the first 5 on-premises activities each month free, for sources behind a corporate firewall. |
Checked November 2026.
What Azure Data Factory does
- More than 90 connectors
- Built-in connectors reach databases, warehouses and SaaS systems, and the page cites throughput of up to 5 Gbps from the connectors, which suits bulk loads between warehouses and storage.
- Code-free data flows
- Intelligent intent-driven mapping lets users design transformations visually, and pipelines can then schedule them, which suits mixed teams of analysts and engineers who share one workspace and one set of pipelines.
- SSIS rehosting
- Existing SSIS packages can be rehosted and extended in Azure, and the vendor says Azure Hybrid Benefit can lower cost by up to 88 percent.
- Git and CI/CD support
- Built-in Git and CI/CD support applies to code-free pipelines too, so changes to pipelines can be reviewed, versioned and promoted between environments like application code, with a history of who changed what.
- Alerts in Azure alert groups
- Alerts appear in Azure alert groups, so failures reach the same notification channels, such as email or incident tools, that operations teams already use for other Azure monitoring signals and resources.
- Concurrency limits by runtime
- Managed VNET pipelines list up to 50 concurrent activities and external pipelines up to 800, which sets how much work can run in parallel before activities start to queue behind each other.
How Azure Data Factory handles your data
The Azure Data Factory product and pricing pages we opened do not state whether customer data trains models, how long data is retained, in which regions it is stored or which certifications apply. They describe a Managed VNET runtime and a Self-Hosted runtime that keeps orchestration on premises. Microsoft publishes compliance and privacy documentation in other places, which was not part of this review. The vendor's policy was not checked for this, so ask your Microsoft account team for the data protection terms and region list before moving regulated pipelines.
Where Azure Data Factory falls short in practice
- Prices not readable. The pricing page showed placeholder values for run rates, so cost cannot be modeled from what we opened and an Azure pricing calculator run is needed before anyone can promise a budget to finance.
- Daily limit on low-frequency activities. Low-frequency activities run no more than once a day, so a job that must run hourly is billed as a regular activity, and the free allowance will not cover it.
- No AI feature on the product page. The page names none. Azure Copilot and Microsoft Foundry appear only in site navigation, and Autonomous ETL is named without being described as AI, so verify scope.
Works with
Alternatives to Azure Data Factory
Azure Data Factory: common questions
How is Azure Data Factory priced?
Orchestration is billed per 1,000 runs, covering activity runs, trigger executions and debug runs, on three runtimes. The captured pricing page showed placeholders instead of numbers, so we cannot quote a rate. The first five low-frequency activities each month are free, and use past 100 activities gets 20 percent off.
Does Azure Data Factory have a free tier?
The pricing page lists the first five low-frequency activities each month as free, both in the cloud and on premises. That is a small allowance, not a full free plan, so budget for runs beyond it. Confirm current rates in the Azure pricing calculator.
Can Azure Data Factory run my SSIS packages?
Yes. The product page says SSIS can be rehosted and extended in Azure, and states savings of up to 88 percent with Azure Hybrid Benefit. That is a vendor claim that depends on your licenses, so compare it against your own license position.
Does Azure Data Factory include AI features?
The product page opened names no AI feature. Azure Copilot and Microsoft Foundry appear only in the site navigation, and the page mentions Autonomous ETL without describing it as AI. If AI authoring matters, ask Microsoft what is available for Data Factory today.
How many connectors does Azure Data Factory have?
The product page says more than 90 built-in connectors, with throughput of up to 5 Gbps. Named sources in the material include Amazon Redshift, Google BigQuery, Salesforce, ServiceNow, Oracle Exadata, Teradata, Azure Synapse Analytics and SQL Server. Check the connector list for your exact source and version.