Free Tier

You can get started with AWS DeepRacer console service for free. Upon first use of AWS DeepRacer simulation in the AWS console, new customers will get 10 hours of Amazon SageMaker training, and 60 Simulation Unit (or SU) hours for Amazon RoboMaker in the form of service credits. The service credits are applied at month end and are available for 30 days. A typical AWS DeepRacer simulation uses up to 6 SUs per hour, thus the free-tier will allow you to run up-to 10 hours for free when running a typical AWS DeepRacer simulation.

AWS DeepRacer integrates with Amazon SageMaker for reinforcement learning model training, AWS RoboMaker to provide the racing simulator, Amazon Kinesis Video Streams for video streaming of virtual simulation footage, Amazon S3 for model storage, and Amazon CloudWatch for log capture. The first time that you use the AWS DeepRacer service, the required IAM roles needed to call the integrated services on your behalf, and the required Virtual Private Cloud (VPC) stack to run Amazon SageMaker and AWS RoboMaker, will be created in your account. There is no cost associated with creating and using a VPC, but standard EC2 data transfer rates apply to the networking traffic incurred by your simulation jobs. For more details about EC2 data transfer rates, click here. A Network Address Translation (NAT) Gateway is created as part of the VPC stack. The NAT Gateway is used to transfer data from your VPC to Amazon S3, Amazon CloudWatch, and Amazon Kinesis Video Streams using private subnets. You will be charged for each hour that the NAT Gateway is provisioned and available, and per gigabyte of data processed through the NAT. You will receive AWS Promotional Credits to offset the cost of the NAT Gateway availability. For more about NAT Gateway pricing, click here.

Pricing example #1 (when outside the Free Tier)

When getting started with AWS DeepRacer, developers need to train a reinforcement learning model. The AWS DeepRacer console service will guide you through creating your first model with a series of sensible defaults, followed by training in the virtual simulator. We recommend that you train your model for at least an hour using the default settings, but training time may change depending on your parameters. During training, AWS DeepRacer will use AWS services on your behalf, including Amazon SageMaker, AWS RoboMaker, and others. For example, AWS RoboMaker creates the simulation environment used during training.

After training completes, you will use the AWS DeepRacer console service to evaluate the performance of your trained model, by allowing it to drive autonomously around the virtual track. This allows you to benchmark the performance of your model. Here, AWS DeepRacer will again use services on your behalf, including AWS RoboMaker, Amazon S3, and Amazon Kinesis Video Streams. The price of these activities is below:

Amazon SageMaker
Service Training Instance Consumed hours Cost per hour Total ($)
Amazon SageMaker - Training ML.C4.2XL 1 $0.56 $0.56
AWS RoboMaker
Service Estimated Simulation Units per Hour Consumed hours Cost per simulation unit per hour ($) Total ($)
AWS RoboMaker - Training 6 1 $0.40 $2.40
AWS RoboMaker - Evaluation 6 0.17 $0.40 $0.40
Total       $2.80
Amazon S3
Service Estimated data (GB) Cost per GB ($) Total ($)
Amazon S3 - Training - Ingest 0.832 $0.00 $0.00
Amazon S3 - Evaluation - Retrieve 0.025 $0.00 $0.00
Total 0.857   $0.00
Amazon Kinesis Video Streams
Service Estimated Data per hour (GB) Hours Total Data (GB) Cost per GB ($) Total ($)
KVS - Training - Store 0.3 1 0.3 $0.0085 $0.00
KVS - Training - HLS stream 0.3 1 0.3 $0.0119 $0.00
KVS - Evaluation - Store 0.3 0.17 0.05 $0.0085 $0.00
KVS - Evaluation - HLS stream 0.3 0.17 0.05 $0.0119 $0.00
Total     0.7   $0.01
Amazon CloudWatch Metrics
Service Estimated metric requests per hour Hours Total metric requests Cost per 1000 metric requests ($) Total ($)
Amazon CloudWatch - Training - Requests 60 1 60 $0.01 $0.00

In this example, the total cost is $3.36. 

Pricing Example #2 (when outside the Free Tier)

While the model you created in Example #1 is a good start, you think you can improve its driving behavior and get a better lap time as a result. You go through 3 iterations of editing your model's parameters, training it, and then evaluating it in the virtual simulator. Your model is now driving much better and the lap time has improved significantly. Finally, you decide to test your optimizations in the real-world by deploying the final model to your AWS DeepRacer car.

Training costs
Service Number of 1 hour jobs Estimated cost per 1 hour training ($) Total
Amazon SageMaker 3 $0.56 $1.67
AWS RoboMaker 3 $2.40 $7.20
Amazon Kinesis Video Streams 3 $0.01 $0.02
Amazon CloudWatch 3 $0.001 $0.002
Total     $8.89
Service Number of 10 minute jobs Estimated cost per 10 min evaluation ($) Total
AWS RoboMaker 3 $0.40 $1.20
Amazon Kinesis Video Streams 3 $0.001 $0.01
Total     $1.21
Model deployment
Service Estimated data (GB) Cost per GB ($) Total ($)
Amazon S3 - Retrieve 0.15 $0.09 $0.01

In this example, the total cost is $10.10.

Pricing Example #3 (when outside the Free Tier)

You decide to leave your models from examples #1 and #2 on Amazon S3 and data in Amazon Kinesis Video Streams. The following table shows the cost for a month of storage.

Service Estimated Data Stored (GB)  Cost per GB per month Total
Amazon S3 - Model Data 2.4 $0.02 $0.08
Amazon Kinesis Video Streams 2.8 $0.02 $0.06
Total 5.2   $0.14

The total storage cost for a month is $0.14.

Additional pricing resources

TCO Calculator

Calculate your total cost of ownership (TCO)

Simple Monthly Calculator

Easily calculator your monthly costs with AWS

Economics Resource Center

Additional resources for switching to AWS

Getting Started

Information on getting started with AWS DeepRacer.

Learn more 
AWS DeepRacer League Rules

AWS Information on AWS DeepRacer League rules and eligibility requirements.

View the Rules 
Pre-order your AWS DeepRacer

Get hands-on with RL, experiment, and learn through autonomous driving.

Pre-Order now