Amazon SageMaker v Overview Features + ML Tools + Pricing FAQs Getting Started ~ Customers Machine Learning » Amazon SageMaker PricingAmazon SageMaker pricing Cc) CO) ‘Amazon SageMaker helps data scientists and developers to prepare, build, train, and deploy high-quality machine learning (ML) models quickly by bringing together a broad set of capabilities purpose-built for ML. SageMaker supports the leading ML frameworks, toolkits, and programming languages. With SageMaker, you pay only for what you use. You have two choices for payment: an On-Demand Pricing that offers no minimum fees and no upfront commitments, and the SageMaker Savings Plans that offer a flexible, usage-based pricing model in exchange for a commitment to a consistent amount of usage Amazon SageMaker Free Tier ‘Amazon SageMaker is free to try. As part of the AWS Free Tier, you can get started with Amazon SageMaker for free. Your free tier starts from the first month when you create your first SageMaker resource. The details of the free tier for Amazon SageMaker are in the table below. ‘Amazon SageMaker capability Studio notebooks, and notebook instances RStudio on SageMaker Data Wrangler Feature Store Training Free Tier usage per month for the first 2 months 250 hours of ml.t3. medium instance on Studio notebooks OR 250 hours of ml.t2 medium instance or ml.t3.medium instance on notebook instances 250 hours of mL:t3.medium instance on RSession app AND free mL.t3.medium instance for RStudioServerPro app 25 hours of mLmS.4xlarge instance 10 million write units, 10 million read units, 25 GB storage (standard online store) 50 hours of m4.xlarge or mS.xlarge instances‘Serverless Inference 150,000 seconds of on-demand inference duration Canvas 160 hours/month for session time HyperPod, 50 hours of m5.xlarge instance AWS Pricing Calculator Calculate your Amazon SageMaker and architecture cost in a single estimate. Create your custom estimate now » On-Demand Pricing Studio Classic ‘Amazon SageMaker Studio Classic Studio Classic offers one-step Jupyter notebooks in our legacy IDE experience. The underlying compute resources are fully elastic and the notebooks can be easily shared with others, allowing seamless collaboration. You are charged for the instance type you choose, based on the duration of use. Region: US East (Ohio) *mLt3.xlarge 4 16 GiB $0.20 mLt3.2xlarge 8 32.GiB $0399 mLms.large 2 8Gi8 $0.115 mlm5.xlarge 4 16 GiB $0.23 mlm5.2xlarge 8 32.GiB $0,461 mlLms.4xlarge 16 64 GiB $0.922 mLms.8xlarge 32 128 GiB $1,843 mlm5.12xlarge 48 192 GiB $2.765 mLms.16xlarge 64 256 GiB $3.686 mlm5.24xlarge 96 384 GiB $5.53 miLmSd.large 2 8GiB $0.136 mimSdxlarge 4 16 GiB $0.271 mlmSd.2xlarge 8 32GB $0542 mimSd4xlarge 16 64 GiB $1,085 mimSd.8xlarge 32 128 GiB $2.17 mlmSd,12xlarge 48 192 GiB $3,254 mLmSd.16xlarge 64 256 GiB $4339 mlmSd.24xlarge 96 384 GiB $6,509 Compute Optimized vcPU Memory Price per Hour mlcS.large 2 4GiB $0.102 mlcS.xlarge 4 8GiB $0,204mlc5.9xlarge ml.c5.12xlarge mLc5.18xlarge Memory Optimized mLrS.large mbr5.xlarge mLrS.2xlarge mLrs.4xlarge mUr5.8xlarge mLr5.12xlarge mLr5.16xlarge mLr5.24xlarge Accelerated Computing mlp3.2xlarge mLp3.8xlarge mLp3.16xlarge ml.g4dn.xlarge mlg4dn.2xlarge mlg4dn.4xlarge ml.g4dn.8xlarge 36 48 72 16 32 48 64 96 vcPU 32 64 16 32 72.6i8 96 GiB 144 GiB Memory 16 GiB 32.6i8 64 GiB 128 GiB 256 GiB 384 GiB 512 GiB 768 GiB Memory 61 GiB 244 GiB 488 GiB 16 GiB 32.618 64 GiB 128 GiB $1.836 $2.448 $3.672 Price per Hour $0.151 $0.302 $0.605 $1.21 $2419 $3.629 $4838 $7.258 Price per Hour $3.825 $14,688 $28,152 $0.7364 $0.94 $1,505 $2.72mltrn1.2xlarge 8 mltrn1.32xlarge 128 mLtrnin.32xlarge 128 Jupytertab Code Editor Rstudio Notebook Instances Processing TensorBoard Data Wrangler Feature Store Training MLflow Real-Time Inference Asynchronous Inference Batch Transform Serverless Inference JumpStart Profiler 32.GiB 512 GiB 512 GiB $1.54 $24.73 $28.50Instance details ‘Amazon SageMaker PS instance product details Instance Size vCPUs Instance GPU GPU Total Memory Network GPUDirect GPU Pei Memory Model GPU perGPU Bandwidth © RDMA _—to Pee! (Tip) memory (GB) (Gbps) (GB) mlp5.48xlarge 192 2 NVIDIA 8 640 80 3200 Yes 900 GB, H100 HEM3, EFAv2 NvSwite SS > ‘Amazon SageMaker P4d instance product details Instance Size vCPUs Instance GPU GPUs Total Memory Network GPUDirect_ GF Memory Model GPU perGPU Bandwidth = RDMA _—Pee (GiB) memory (GB) (Gbps) Pe (GB) mlp4d.24xlarge 96 = 1152, NVIDIA 8 = 320 40 400ENA Yes 600¢ ‘A100 HBM 2 ‘AND EFA Nvsv mlp4de.24xlarge 96 1152, NVIDIA 8 = 640 80 400ENA Yes 600¢ A100 HNM2e and EFA Nvsv LS > Amazon SageMaker P3 instance product details Instance Size vCPUs Instance GPU GPUs Total Memory Network © GPU _Instanci Memory Model GPU perGPU Bandwidth Peer —_ Storage (GiB) (GB) (Gbps) (cB)mlp3.8xlarge «32,244 NVIDIA 4 64 16 10 NVLink EBS-Onl vi00 mLp3.16xlarge 64 488 NVIDIA 8 128 16 25 NVLink EBS-Onl v100 mlp3dn.24xlarge 96 768 NVIDIA 8 256 32 100 © NVLink 2x 900 vi00 NVMess ‘Amazon SageMaker P2 instance product details Instance Size vCPUs Instance GPU GPUs Total += Memory ~—_—Network EBS Memory Model GPU per GPU Bandwidth Bandwidth (GiB) memory (GB) (Gbps) (Gbps) (GB) mlp2xlarge 4 6 NVIDIA 1 12 2 Up to 10, High k80 mlp28xlarge 32 488 NVIDIA 8. 96 2 10 10 K80 mlp2.16xlarge 64 732 NVIDIA 16 192 2 25 20 Keo Amazon SageMaker G4 instance product details Instance Size vCPUs Instance GPU GPUs Total Memory Network Instance —EBS Memory Model GPU per GPU Bandwidth Storage Bandw (GiB) memory (GB) (Gbps) (6B) (Gbp (GB) mlg4dnxlarge 4 16 NVIDIA 1 16 16 Upto25-1x125 Upto 14 NVMe sD4 mlg4dngxlarge 32 128 NVIDIA T4 mlg4dn.t6xlarge 64-256 NVIDIA 4 mlg4dn.12xlarge 48192 NVIDIA 14 ‘Amazon SageMaker G5 instance product details Instance Size vCPUs Instance GPU GPUs. Memory Model (Gi) mlgSnxlarge 4 16 NVIDIA A106 mlg5.2xlarge 8 32 NVIDIA A106 mlg5.4xlarge 16 64 NVIDIA A106 mlg5.8xlarge 32 128 NVIDIA A106 mlgS.16xlarge 64 «256 NVIDIA A106 mlg5.12xlarge 48 192 NVIDIA A106 16 16 50 16 16 50 64 16 50 Total Memory — Network GPU per GPU Bandwidth Memory (GB) (Gbps) (GB) 24 24 Upto 10 24 24 Upto 10 24 24 Up to 25, 24 24 25 24 24 2s 96 24 40 NVMe ss 1x900 95 NVMe ssD 1x900 95 NVMe ssD 1x900 95 NVMe ssD EBS Instance Bandwidth Storag (Gbps) (GB) Upto3.5 — 1x250 Upto3.5 — 1x45¢ 8 1600 16 1900 16 1x1901 16 113801Amazon SageMaker Trn1 instance product details Instance Size mLtrn1.2xlarge mltrn1.32xlarge vCPUs 128 Memory (GiB) 32 512 Accelerators 16 ‘Amazon SageMaker Inf1 instance product details Total Memory Instance Accelerator —per_— ‘Storage Memory Accelerator (GB) (GB) (GB) 32 32 1x 500 NvMe ssD 512 32 4x 2000 NVMe ssD Network I Bandwidth Bani (bps) (6 Upto125 Up 800 Instance Size mLinf1 xlarge mlinf1.2xlarge mLinf1.6xlarge mLinf1.24xlarge vCPUs 24 96 Memory (GiB) 16 48 192 Inferentia Accelerators 16 Total Accelerator Memory (cB) 8 32 128 Memory per Accelerator (GB) 8 —_—_—_ Instance Storage EBS only EBs only EBs only EBS only Inte accele! Interco Ni NA Ye yeMemory Accelerator Interco (GB) (GB) mlinf2.xlarge 4 16 1 32 32 EBS NA only mlinf2.8xlarge 32 128 1 32 32 EBS Ni only mlinf2.24xlarge 96 384 6 196 32 EBS Ye only mlinf2.48xlarge 192 768 2 384 32 EBS Ye only SS > Amazon SageMaker Studio ‘Amazon SageMaker Studio is a single web-based interface for complete ML development, offering a choice of fully managed integrated development environments (IDEs) and purpose-built tools. You can access SageMaker Studio free of charge. You are only charged for the underlying compute and storage that you use for different IDEs and ML tools within SageMaker Studio. You can use many services from SageMaker Studio, AWS SDK for Python (Boto3), or AWS Command Line Interface (AWS CLI) including the following + IDEs on SageMaker Studio to perform complete ML development with a broad set of fully managed IDEs, including JupyterLab, Code Editor based on Code-OSS (Visual Studio Code - Open Source), and RStudio * SageMaker Pipelines to automate and manage ML workflows + SageMaker Autopilot to automatically create ML models with full visibility + SageMaker Experiments to organize and track your training jobs and versions * SageMaker Debugger to debug anomalies during training * SageMzker Model Monitor to maintain high-quality models + SageMaker Clarify to better explain your ML models and detect bias + SageMaker JumpStart to easily deploy ML solutions for many use cases. You may incur charges from other ‘AWS services used in the solution for the underlying API calls made by Amazon SageMaker on your behalf.To use Amazon 0 Developer in JupyterLab, you must subscribe to Amazon Q Developer Pro, Amazon Q Developer pricing is available here. Foundation model evaluations SageMaker Clarify supports foundation model evaluations with both automatic and human-based evaluation methods. Each of these has different pricing. If you are evaluating a foundation model from Amazon SageMaker JumpStart that is not yet deployed to your account, SageMaker will temporarily deploy the JumpStart model on a SageMaker instance for the duration of the inference. The specific instance will conform to the instance recommendation provided by JumpStart for that model. Automatic evaluation: Foundation model evaluations run as SageMaker processing job. The evaluation job will invoke SageMaker Inference. Customers are charged for the inference and for the evaluation job. Customers are charged only for the duration of the evaluation job. The cost of the evaluation job would be the sum of the cost per hour of the evaluation instance and the sum of the cost per hour of the hosting instance. Human-based evaluation: ‘When you use the human-based evaluation feature where you bring your own workforce, you are charged for three items: 1) SageMaker instance used for inference, 2) the instance used to run the SageMaker Processing Job that hosts the human evaluation, and 3) a charge of $0.21 per completed human evaluation task. A human task is defined as an occurrence of a human worker submitting an evaluation of a single prompt and its associated inference responses in the human evaluation user interface. The price is the same whether you have 1 or 2 models in your evaluation job or you bring your own inference and also the same regardless of how many evaluation dimensions and rating methods you include. The $0.21 per task pricing is the same for all AWS regions. There is no separate charge for the workforce, as the workforce is supplied by you. AWS- managed evaluation: For an AWS-managed expert evaluation, pricing is customized for your evaluation needs in a private engagement while working with the AWS expert evaluations team. Amazon SageMaker Studio Lab You can build and train ML models u: developers, academics, and data scientists a no-configuration development environment to learn and experiment. with ML at no additional charge. jing Amazon SageMaker Studio Lab for free. SageMaker Studio Lab offers Amazon SageMaker Canvas‘Amazon SageMaker Data Labeling provides two data labeling offerings, Amazon SageMaker Ground Truth Plus and Amazon SageMaker Ground Truth. You can learn more about Amazon SageMaker Data Labeling, a fully ing datasets for ML. managed data labeling service that makes it easy to build highly accurate tr Amazon SageMaker shadow testing ‘SageMaker helps you run shadow tests to evaluate a new ML model before production release by testing its performance against the currently deployed model. There is no additional charge for SageMaker shadow testing other than usage charges for the ML instances and ML storage provisioned to host the shadow model. The pricing for ML instances and ML storage dimensions is the same as the real-time inference option specified in the preceding pricing table, There is no additional charge for data processed in and out of shadow deployments. Amazon SageMaker Edge Learn more about pricing for Amazon SageMaker Edge to optimize, run, and monitor ML models on fleets of edge devices. Amazon SageMaker Savings Plans ‘Amazon SageMaker Savings Plans help to reduce your costs by up to 64%, The plans automatically apply to eligible SageMaker ML instance usage, including SageMaker Studio notebooks, SageMaker notebook instances, SageMaker Processing, SageMaker Data Wrangler, SageMaker Training, SageMaker Real-Time Inference, and SageMaker Batch Transform regardless of instance family, size, or Region. For example, you can change usage from a CPU instance ml.cS.xtarge running in US East (Ohio) to a mL.nf1 instance in US West (Oregon) for inference workloads at any time and automatically continue to pay the Savings Plans price. Learn more » Total cost of ownership (TCO) with Amazon SageMaker ‘Amazon SageMaker offers at least 54% lower total cost of ownership (TCO) over a three-year period compared to other cloud-based self-managed solutions. Learn more with the complete TCO analysis for Amazon SageMaker.Pricing example #1: JupyterLab Pricing example #2: Code Editor Pricing example #3: Studio Classic Pricing example #4: RStudio ing example #5: Processing ing example #6: Data Wrangler ing example #7: Feature Store ing example #8: Training ing example #9: MLflow ing example #10: Real-time inference @ OOOO OO 08 © ©ing example #12: Batch Transform ing example #13: On-demand Serverless Inference Pri Inference ing example #14: Provisioned Concurrency on Serverless ing example #15: Jumpstart ing example #16: HyperPod Cluster Pricing Example #17: Foundation model evaluations (automatic evaluation) Pricing Example #18: Foundation model evaluations (human- based evaluation) Next steps Oo © © O © © ©Console Sign In to the Console Learn About AWS What Is AWS? What Is Cloud Computing? AWS Accessibility Resources for AWS Getting Started Training and Certification AWS Solutions Library Architecture Center Product and Technical FAQs Analyst Reports Developers on AWS Developer Center SDKs & Tools NET on AWS Python on AWS Java on AWS PHP on AWSWhat is Generative AI? AWS Cloud Security What's New Blogs Press Releases Help Contact Us Get Expert Help File a Support Ticket AWS re:Post Knowledge Center AWS Support Overview Legal AWS Careers C Create an AWS Account +) f in © Da@s ‘Amazon is an Equal Opportunity Employer: Minority / Women / Disability / Veteran / Gender Identity / Sexual Orientation / Age. Language wel Bahasa Indonesia | Deutsch | English | Espajiol | Francais | Italiano |Privacy | Accessibility | Site Terms Cookie Preferences © 2024, Amazon Web Services, Inc. or its affiliates. All rights reserved.