New enhancements in Private Channels in Microsoft Teams unlock their full potential
August 2, 2025[Launched] Generally Available: Azure File Sync Arc Extension
August 2, 2025
At Microsoft Build 2025, we announced Azure SRE Agent, an AI tool designed to streamline incident response, improve service uptime, and cut operational costs. The SRE Agent uses AI Models in Azure AI Foundry to quickly analyze logs and metrics for root cause analysis and issue resolution. This advanced AI agent enhances incident and infrastructure management in Azure, allowing site reliability engineers (SRE) to focus on higher-value tasks.
Starting September 1, 2025, billing will begin for Azure SRE Agent. Please visit the Azure pricing page to see the pricing in your region. This article describes the billing model and example scenarios.
Overview of Azure SRE Agent
Azure SRE Agent is a powerful tool that uses machine learning and advanced monitoring capabilities to proactively manage cloud resources, detect anomalies, and execute remediation tasks. The agent’s unique dual-action architecture allows it to remain always-on, learning from application behavior, and ready to spring into action the moment intervention is required. Whether you’re running small-scale workloads or managing complex, mission-critical applications, Azure SRE Agent brings automation, reliability, and peace of mind to your operations.
As part of the broader Agentic DevOps evolution, SRE agent reimagines cloud operations by embedding intelligent agents across the software development lifecycle. This AI-native approach transforms DevOps from manual coordination to intelligent orchestration, enabling faster, more secure, and scalable cloud-native application development.
Introducing a new billing unit: Azure Agent Unit
Billing for Azure SRE Agent is based on Azure Agent Units (AAU). AAUs standardize measuring agentic processing across all pre-built Azure agents. This common metric makes it easier for you to adopt new agents, compare different agents, and understand pricing in a transparent, predictable way.
Scalable usage-based billing model
The Azure SRE Agent billing model includes two components: always-on flow for continuous monitoring and active flow for incident mitigation and other tasks where the AI is actively engaged. Total cost is a combination of the fixed baseline cost and incremental usage-based costs.
- Each agent is billed at 4 Azure Agent Units per hour as the baseline cost for always-on flow.
- Each task the agent executes is billed at an additional 0.25 Azure Agent Units per second as usage-based cost for active flow.
Let us explore these concepts below.
1.Baseline, fixed cost component: Always-on flow
Once created, the agent is continuously running in the background. It monitors telemetry, learns from normal behavior, and is primed to respond instantly. This foundational layer provides the assurance that your environments are safeguarded by an intelligent agent 24/7, no matter the scale or complexity.
This model gives you a predictable baseline cost associated with an active agent. With the always-on flow the agent is always learning so that it is always ready to respond to issues as they happen.
2. Usage based variable cost component: Active flow
When the agent detects a situation requiring attention—a system anomaly, a failure, or a defined threshold breach—it activates the AI components. During this time, the agent carries out discrete, measurable tasks, such as automated incident response, resource scaling, or system recovery.
This usage-based component allows organizations to pay precisely for what they use, scaling costs in line with actual operational demand.
This table summarizes the billing model:
|
Billing component |
Type |
Pay as you go rate |
|
Baseline (fixed cost) |
Always-on flow |
4 Azure Agent Units (AAU) per agent per hour |
|
Usage-based (variable cost) |
Active flow |
0.25 Azure Agent Units (AAU) per agent task per second |
Total cost is a combination of the fixed baseline cost and incremental usage-based costs. Please visit the Azure pricing page for price in your region.
Example Scenarios
To illustrate Azure SRE Agent’s pricing, consider these examples. Please use the Azure price calculator to estimate your hourly or monthly costs for using Azure. In these examples, we assume unit price of 1 Azure Agent Unit is $0.10, and there are 730 hours (31 days) in a month:
|
Scenario description |
A. Baseline (fixed) cost estimate |
B. Usage-based (variable) cost estimate |
(A+B) Total cost estimate |
|
Minimal incidents: Over the course of a month, the development team keeps Azure SRE Agent running across several test environments. No automated incidents occur. The team interacts with the agent four times, each session creating 5 minutes of work for the agent. Waiting time, where the agent is waiting for human response, is not counted as active flow. |
Always-on flow: 4 AAU X 730 hours X $0.10 per AAU = $292 per month per agent |
Active flow: 0.25 AAU X 4 Active tasks per month X 300 seconds runtime on each task X $0.10 per AAU = $30 per month per agent |
Total cost: $322 per month per agent |
|
High operational load: An enterprise has deployed five Azure SRE Agents across five different production workloads. These agents are integrated with the incident management system, telemetry stores, and communication tools to manage all customer-reported incidents. On average, each agent handles two customer-reported incidents per day, requiring approximately 10 minutes of execution time to analyze multiple data sources, identify the root cause, and apply the remediation runbook.
|
Always-on flow: 4 AAU X 730 hours X $0.10 per AAU = $292 per month per agent
$1460 per month for five agents |
Active flow: 0.25 AAU X 2 Active tasks per day x Tasks run for 31 days a month x 600 seconds runtime on each task X $0.10 per AAU = $930 per month per agent
$4650 per month for five agents |
Total cost: $1222 per month per agent
$6110 per month for five agents |
|
Burst events: An ISV deploys one Azure SRE Agent to manage five large workloads across five different customers. While most of the month passes without incidents, a major weekend outage affects all five customers simultaneously, triggering 50 incidents. Each incident takes about 5 minutes to diagnose and resolve. The Azure SRE Agent handles these incidents in parallel. |
Always-on flow: 4 AAU X 730 hours X $0.10 per AAU = $292 per month per agent |
Active flow: 0.25 AAU X 50 Active tasks per day X Tasks run for 1 day a month X 300 seconds runtime on each task X $0.10 per AAU = $375 per month per agent |
Total cost: $667 per month per agent |
Note that these calculations are for illustrative purposes only.
Getting Started
If you are interested in deploying Azure SRE Agent, please join the wait list. We are rolling out this service in phases, and you will hear from the team as soon as we are ready to onboard you. During the preview period, features and pricing details are subject to change ahead of the product’s general availability.
Here are additional resources to help you get ready to fully explore the agent’s capabilities.