Azure Logic Apps Automation pricing
Azure Logic Apps Automation pricing
Section titled “Azure Logic Apps Automation pricing”Logic Apps Automation is Azure’s fully managed platform for business process automation and orchestration, including AI-powered automation.
Logic Apps Automation uses consumption-based pricing, so you pay based on the resources and capabilities that your automations use. The billing meters fall into two groups: Core Platform meters that apply to every workflow, and Optional Capabilities that are billed only when a workflow uses them.
[!IMPORTANT] Azure Logic Apps Automation is currently in public preview. Preview pricing and availability are subject to change before general availability.
[!NOTE] The prices shown on this page are for the East US region. Azure prices can vary by region. To view current pricing for your region and currency, see Pricing - Logic Apps | Microsoft Azure.
Core platform pricing
Section titled “Core platform pricing”Core Platform meters cover the managed environment, workflow execution, and run-history storage required to host and run your automations.
| Component | Meter | Price | How pricing works |
|---|---|---|---|
| Managed environment | Environment hour | $0.042 per environment hour | Charged for each hour that a managed, customer-isolated environment is provisioned. |
| Workflow runtime | Execution-second | $0.00008 per execution-second | This is the core execution time per second for an Automation application. The default profile includes 1 vCPU of core compute and 2 GiB of memory. The runtime host is also active while you author workflows in the portal. |
| Data retention | GB per month | $0.12 per GB per month | Charged for storage used by workflow run history. Run history is retained for up to approximately 90 days. |
Optional capability pricing
Section titled “Optional capability pricing”Optional Capabilities extend a workflow with connectors, AI, managed knowledge, or isolated code execution. You incur these charges only when your automation uses the corresponding capability.
| Component | Meter | Price | How pricing works |
|---|---|---|---|
| Standard managed connectors | Connector action | $0.000125 per action | Charged each time a workflow runs an action through a Standard managed connector. |
| Enterprise managed connectors | Connector action | $0.001 per action | Charged for Enterprise connector operations, including knowledge read and write transactions. |
| AI usage | Azure Agent Credits (AACs) | Charged as AACs | AI usage is converted to AACs based on the selected model and token consumption for AI-assisted authoring, agent reasoning, inference, embeddings, and other AI operations. See Azure OpenAI Service pricing. |
| Knowledge storage | GB per month | 33 Azure Agent Credits (AACs) per GB per month, equivalent to $0.33 per GB per month | Charged for storage used by the managed knowledge layer. Embeddings, inference, and transactions are billed separately through AI usage and connector actions. |
| Sandbox: isolated code execution | vCPU-second and GiB-second | $0.000024 per vCPU-second and $0.000003 per GiB-second | Charged for the compute and memory consumed while custom code runs in an isolated sandbox. |
Public preview regions
Section titled “Public preview regions”Azure Logic Apps Automation is available in the following Azure regions during public preview:
- Australia East
- East Asia
- North Central US
- Southeast Asia
- Sweden Central
- UK South
- West US
Region availability determines where you can create an Automation project. Azure OpenAI model availability and quota can also vary by region and subscription.
Example monthly cost
Section titled “Example monthly cost”The following example uses East US prices to show how charges from different meters contribute to the estimated monthly cost. The usage figures are illustrative and do not represent a recommended configuration or a typical customer workload.
Assume an organization has the following monthly usage:
- One managed environment provisioned for 730 hours.
- The Automation application consumes 150,000 execution-seconds. This includes time used to execute workflows and time when the runtime host is active during workflow authoring in the portal.
- 1,000 Standard connector actions.
- 100 Enterprise connector actions generated by knowledge read and write operations.
- 20 GB of retained workflow run history.
- 5 GB of knowledge storage.
- 100 sandbox executions. Each execution uses one vCPU and 2 GiB of memory for 10 seconds.
- AI usage, charged as AACs based on the selected model and token consumption.
| Meter | Calculation | Estimated cost |
|---|---|---|
| Managed environment | 730 hours × $0.042 | $30.66 |
| Automation application execution time | 150,000 execution-seconds × $0.00008 | $12.00 |
| Standard connector actions | 1,000 actions × $0.000125 | $0.125 |
| Enterprise connector actions for knowledge operations | 100 actions × $0.001 | $0.10 |
| Data retention | 20 GB × $0.12 | $2.40 |
| Knowledge storage | 5 GB × $0.33 | $1.65 |
| Sandbox vCPU | 100 executions × 10 seconds × 1 vCPU × $0.000024 | $0.024 |
| Sandbox memory | 100 executions × 10 seconds × 2 GiB × $0.000003 | $0.006 |
| AI usage | Model and token consumption converted to AACs | Additional AAC charges |
| Estimated monthly total, excluding AI usage | $46.97 |
The final bill can be higher or lower depending on the Automation application’s total execution time, retained data, connector activity, AI usage, and sandbox resource consumption. For AI rates, see Azure OpenAI Service pricing.
Pricing notes
Section titled “Pricing notes”- Pricing changes are being rolled out in phases across Azure regions. Pricing might not be available in every region at the same time.
- AI tokens are consumed by Agents, the Chat Assistant, and Knowledge. If you bring your own model, AI token usage is not charged through Logic Apps Automation.
- The prices shown are East US prices in US dollars and are estimates only. Prices can vary by region, agreement, purchase date, currency exchange rate, and applicable taxes. Check Azure Logic Apps pricing for current regional pricing.
- Knowledge storage pricing does not include AI processing or connector transactions.
- AI usage is charged as AACs. The number of AACs consumed depends on the selected model and the input and output tokens processed.