AWS IoT Things Graph makes it easy to rapidly build IoT applications by connecting devices and web services from different vendors that speak different protocols. Using the visual editor in AWS IoT Things Graph, you can wire together devices and web services as a series of steps called a flow. Once a flow is created, you can either run it in the AWS Cloud or deploy it to your AWS IoT Greengrass-enabled device.
With AWS IoT Things Graph, you pay only for what you use with no minimum fees or mandatory service usage. You are billed separately depending on whether you run your application in the AWS Cloud or at the edge on an AWS Greengrass Core.
Running flows in the cloud or on the edge
Free Tier for AWS Cloud
The AWS Free Tier offers new customers the first 5,000 steps per month free during the first twelve months. Usage beyond these levels is billed at the published rates.
Steps executed in the cloud
When running flows in the cloud, you pay for the number of steps executed in your flow. You are charged per step executed above the Free Tier. See Cloud and Edge Flows Pricing Table for details.
Deployments at the edge
You are charged only when you deploy a new or updated flow to AWS IoT Greengrass devices. You can run flows on an AWS IoT Greengrass device at no cost. A flow deployed to one AWS IoT Greengrass device is metered as one deployment. See Cloud and Edge Flows Pricing Table for details.
Cloud and Edge Flows: Pricing Table
Cloud Deployment: You are charged separately for messages that your devices exchange with AWS IoT Things Graph, billed at AWS IoT Core messaging rates. (Please see the AWS IoT Core pricing page).
Edge Deployment: There are separate charges for AWS IoT Greengrass. Please see the AWS IoT Greengrass pricing page for details. You are also charged separately for messages that your IoT Greengrass devices exchange with the AWS cloud. You are billed at AWS IoT Core messaging rates. Please see the AWS IoT Core pricing page for details.
In this example, we explain how you are charged for a face detection application. The application flow consists of one trigger (a motion sensor) and up to three steps (Camera, Amazon Rekognition, and Screen). The flow is triggered when the motion sensor detects motion (trigger), upon which the Camera takes and sends a picture to Amazon Rekognition for face detection (step 1). Amazon Rekognition analyzes the picture to detect a face (step 2), and if it does, the picture is sent to a Screen for display (step 3).
Note that AWS IoT Things Graph automatically generates editable event labels for connections between two nodes (like visible in the above diagram). These are not counted as steps.
70% of the time, the Amazon Rekognition detects a face in the picture, which results in the execution of all 3 steps: Camera, Rekognition, and Screen. 30% of the time, a face is not detected, which results in only 2 steps being completed: Camera and Amazon Rekognition.
You can choose to run this face detection application only in the AWS Cloud (charged per step), only to your AWS IoT Greengrass-enabled devices (charged per deployment), or both. Pricing examples of a Cloud and edge deployment are below.
Cloud Deployment Option
You deploy 100 flows to the AWS Cloud to manage access at 100 locations in the US East (Northern Virginia) region. Assuming that the motion sensor triggers the flow 100 times every day per location for 30 days, your AWS IoT Things Graph charges would be:
- Total flow executions = 100 flows * 100 executions/flow-day * 30 days = 300,000 flow executions
- Valid path steps executed = 70% * 300,000 flow executions * 3 steps/flow execution = 630,000 steps
- Invalid path steps executed = 30% * 300,000 flow executions * 2 steps/flow execution = 180,000 steps
- Total steps executed = 630,000 steps + 180,000 steps = 810,000 steps
- Total step charges for the month = 810,000 steps * $0.000045/step = $36.45
Edge Deployment Option
You can alternately deploy the face detection workflow across your edge devices. You deploy the above flow to 100 edge gateways running AWS IoT Greengrass at 100 locations (i.e. 1 edge gateway per location) also within the US East (Northern Virginia) region. Your AWS IoT Things Graph charges would be:
- Number of deployments = 1 deployment/location * 100 locations = 100 deployments
- Total deployment charges = 100 deployments * $0.0095/deployment = $0.95
Explore more details about models, workflows, and application performance monitoring
Instantly get access to the AWS Free Tier and get started with AWS IoT Things Graph
Review FAQs about AWS IoT Things Graph