Today, we are thrilled to reveal that DeepSeek R1 distilled Llama and Qwen models are available through Amazon Bedrock Marketplace and Amazon SageMaker JumpStart. With this launch, you can now release DeepSeek AI’s first-generation frontier design, DeepSeek-R1, in addition to the distilled variations varying from 1.5 to 70 billion criteria to construct, experiment, and properly scale your generative AI concepts on AWS.

In this post, we show how to begin with DeepSeek-R1 on Amazon Bedrock Marketplace and SageMaker JumpStart. You can follow similar steps to release the distilled variations of the designs also.

Overview of DeepSeek-R1

DeepSeek-R1 is a big language design (LLM) developed by DeepSeek AI that uses support learning to boost thinking capabilities through a multi-stage training process from a DeepSeek-V3-Base foundation. An essential differentiating feature is its support learning (RL) step, which was utilized to refine the design’s responses beyond the standard pre-training and fine-tuning process. By integrating RL, DeepSeek-R1 can adapt more efficiently to user feedback and objectives, ultimately improving both importance and clarity. In addition, DeepSeek-R1 uses a chain-of-thought (CoT) method, implying it’s geared up to break down complicated inquiries and factor through them in a detailed way. This guided reasoning process allows the design to produce more accurate, transparent, and detailed answers. This design integrates RL-based fine-tuning with CoT capabilities, aiming to produce structured reactions while focusing on interpretability and user interaction. With its comprehensive abilities DeepSeek-R1 has actually caught the market’s attention as a versatile text-generation model that can be integrated into various workflows such as agents, sensible thinking and data analysis jobs.

DeepSeek-R1 utilizes a Mix of Experts (MoE) architecture and is 671 billion specifications in size. The MoE architecture enables activation of 37 billion criteria, enabling efficient inference by routing questions to the most appropriate expert “clusters.” This approach allows the design to concentrate on different issue domains while maintaining overall performance. DeepSeek-R1 requires at least 800 GB of HBM memory in FP8 format for inference. In this post, we will utilize an ml.p5e.48 xlarge instance to deploy the design. ml.p5e.48 xlarge features 8 Nvidia H200 GPUs offering 1128 GB of GPU memory.

DeepSeek-R1 distilled models bring the reasoning capabilities of the main R1 model to more effective architectures based upon popular open designs like Qwen (1.5 B, higgledy-piggledy.xyz 7B, 14B, and 32B) and Llama (8B and 70B). Distillation refers to a process of training smaller sized, more efficient models to imitate the behavior and thinking patterns of the bigger DeepSeek-R1 model, utilizing it as a teacher model.

You can deploy DeepSeek-R1 design either through SageMaker JumpStart or Bedrock Marketplace. Because DeepSeek-R1 is an emerging model, we recommend deploying this model with guardrails in place. In this blog, we will utilize Amazon Bedrock Guardrails to introduce safeguards, prevent harmful material, and examine models against key security criteria. At the time of writing this blog, for DeepSeek-R1 implementations on SageMaker JumpStart and Bedrock Marketplace, Bedrock Guardrails supports just the ApplyGuardrail API. You can create several guardrails tailored to different use cases and use them to the DeepSeek-R1 model, improving user experiences and standardizing security controls across your generative AI applications.

Prerequisites

To release the DeepSeek-R1 model, you need access to an ml.p5e instance. To check if you have quotas for P5e, open the Service Quotas console and under AWS Services, pick Amazon SageMaker, and confirm you’re utilizing ml.p5e.48 xlarge for endpoint use. Make certain that you have at least one ml.P5e.48 xlarge instance in the AWS Region you are releasing. To ask for a limitation boost, develop a limitation increase request and reach out to your account group.

Because you will be releasing this design with Amazon Bedrock Guardrails, make certain you have the proper AWS Identity and Gain Access To Management (IAM) approvals to utilize Amazon Bedrock Guardrails. For guidelines, see Establish authorizations to utilize guardrails for content filtering.

Implementing guardrails with the ApplyGuardrail API

Amazon Bedrock Guardrails allows you to present safeguards, avoid harmful material, and evaluate designs against crucial security criteria. You can execute precaution for the DeepSeek-R1 design using the Amazon Bedrock ApplyGuardrail API. This permits you to apply guardrails to evaluate user inputs and model reactions released on Amazon Bedrock Marketplace and SageMaker JumpStart. You can produce a guardrail using the Amazon Bedrock console or the API. For the example code to create the guardrail, see the GitHub repo.

The basic circulation involves the following actions: First, the system receives an input for the design. This input is then processed through the ApplyGuardrail API. If the input passes the guardrail check, it’s sent to the model for inference. After receiving the model’s output, another guardrail check is applied. If the output passes this final check, it’s returned as the final outcome. However, if either the input or output is intervened by the guardrail, a message is returned indicating the nature of the intervention and whether it took place at the input or output phase. The examples showcased in the following sections show inference using this API.

Deploy DeepSeek-R1 in Amazon Bedrock Marketplace

Amazon Bedrock Marketplace provides you access to over 100 popular, emerging, and specialized foundation models (FMs) through Amazon Bedrock. To gain access to DeepSeek-R1 in Amazon Bedrock, complete the following actions:

1. On the Amazon Bedrock console, choose Model brochure under Foundation models in the navigation pane. At the time of composing this post, you can use the InvokeModel API to conjure up the model. It doesn’t support Converse APIs and other Amazon Bedrock tooling.

1. Filter for DeepSeek as a supplier and pick the DeepSeek-R1 model.
   
   The design detail page supplies necessary details about the model’s capabilities, pricing structure, and execution standards. You can discover detailed usage guidelines, including sample API calls and code bits for integration. The model supports different text generation tasks, consisting of material production, code generation, and concern answering, utilizing its support discovering optimization and CoT reasoning capabilities. The page likewise consists of implementation options and licensing details to help you begin with DeepSeek-R1 in your applications.
2. To start using DeepSeek-R1, select Deploy.
   
   You will be triggered to set up the release details for DeepSeek-R1. The model ID will be pre-populated.
3. For Endpoint name, go into an endpoint name (between 1-50 alphanumeric characters).
4. For Number of circumstances, enter a variety of instances (in between 1-100).
5. For example type, select your instance type. For optimal efficiency with DeepSeek-R1, a GPU-based circumstances type like ml.p5e.48 xlarge is recommended. Optionally, you can configure advanced security and infrastructure settings, including virtual private cloud (VPC) networking, service function permissions, and encryption settings. For many utilize cases, the default settings will work well. However, for production releases, you might desire to review these settings to line up with your company’s security and compliance requirements.
6. Choose Deploy to start utilizing the design.
   
   When the implementation is complete, you can check DeepSeek-R1’s capabilities straight in the Amazon Bedrock playground.
7. Choose Open in playground to access an interactive interface where you can experiment with various triggers and adjust model specifications like temperature and optimum length. When utilizing R1 with Bedrock’s InvokeModel and Playground Console, utilize DeepSeek’s chat template for optimal outcomes. For instance, material for inference.
   
   This is an exceptional method to explore the model’s thinking and text generation capabilities before incorporating it into your applications. The play ground offers immediate feedback, assisting you comprehend how the model reacts to different inputs and letting you tweak your prompts for optimum results.
   
   You can quickly test the design in the play area through the UI. However, to invoke the released design programmatically with any Amazon Bedrock APIs, you require to get the endpoint ARN.
   
   Run inference using guardrails with the released DeepSeek-R1 endpoint
   
   The following code example shows how to perform inference utilizing a released DeepSeek-R1 design through Amazon Bedrock using the invoke_model and ApplyGuardrail API. You can create a guardrail utilizing the Amazon Bedrock console or the API. For the example code to create the guardrail, see the GitHub repo. After you have created the guardrail, use the following code to implement guardrails. The script initializes the bedrock_runtime client, sets up inference criteria, and sends a request to create text based upon a user prompt.
   
   Deploy DeepSeek-R1 with SageMaker JumpStart
   
   SageMaker JumpStart is an artificial intelligence (ML) hub with FMs, built-in algorithms, and prebuilt ML services that you can release with simply a few clicks. With SageMaker JumpStart, you can tailor pre-trained models to your use case, with your information, and deploy them into production utilizing either the UI or SDK.
   
   Deploying DeepSeek-R1 design through SageMaker JumpStart uses two practical approaches: utilizing the user-friendly SageMaker JumpStart UI or executing programmatically through the SageMaker Python SDK. Let’s check out both methods to assist you pick the technique that best fits your needs.
   
   Deploy DeepSeek-R1 through SageMaker JumpStart UI
   
   Complete the following actions to release DeepSeek-R1 utilizing SageMaker JumpStart:
   
   1. On the SageMaker console, choose Studio in the navigation pane.
8. First-time users will be triggered to create a domain.
9. On the SageMaker Studio console, choose JumpStart in the navigation pane.
   
   The model web browser shows available designs, with details like the company name and model abilities.
   
   4. Look for DeepSeek-R1 to view the DeepSeek-R1 model card. Each design card reveals essential details, including:
   
   - Model name - Provider name
10. Task classification (for example, Text Generation). Bedrock Ready badge (if appropriate), suggesting that this model can be signed up with Amazon Bedrock, enabling you to use Amazon Bedrock APIs to invoke the model
    
    5. Choose the design card to view the design details page.
    
    The page includes the following details:
    
    - The design name and provider details. Deploy button to release the design. About and Notebooks tabs with detailed details
    
    The About tab consists of crucial details, such as:
    
    - Model description.
11. License details.
12. Technical specifications.
13. Usage guidelines
    
    Before you release the design, it’s recommended to evaluate the model details and license terms to confirm compatibility with your usage case.
    
    6. Choose Deploy to continue with release.
    
    7. For Endpoint name, utilize the immediately produced name or produce a custom-made one.
14. For example type ¸ pick a circumstances type (default: ml.p5e.48 xlarge).
15. For Initial instance count, enter the number of instances (default: 1). Selecting suitable circumstances types and counts is crucial for expense and efficiency optimization. Monitor your release to adjust these settings as needed.Under Inference type, Real-time inference is selected by default. This is enhanced for sustained traffic and low latency.
16. Review all configurations for precision. For this design, we highly recommend sticking to SageMaker JumpStart default settings and making certain that network isolation remains in location.
17. Choose Deploy to release the model.
    
    The release procedure can take a number of minutes to finish.
    
    When release is total, your endpoint status will change to InService. At this point, the design is ready to accept reasoning requests through the endpoint. You can keep an eye on the release progress on the SageMaker console Endpoints page, which will display appropriate metrics and status details. When the deployment is total, you can invoke the design utilizing a SageMaker runtime client and integrate it with your applications.
    
    Deploy DeepSeek-R1 using the SageMaker Python SDK
    
    To get begun with DeepSeek-R1 using the SageMaker Python SDK, you will need to set up the SageMaker Python SDK and make certain you have the essential AWS consents and environment setup. The following is a detailed code example that demonstrates how to release and utilize DeepSeek-R1 for inference programmatically. The code for releasing the design is provided in the Github here. You can clone the note pad and range from SageMaker Studio.
    
    You can run extra demands against the predictor:
    
    Implement guardrails and run reasoning with your SageMaker JumpStart predictor
    
    Similar to Amazon Bedrock, you can likewise use the ApplyGuardrail API with your SageMaker JumpStart predictor. You can create a guardrail utilizing the Amazon Bedrock console or the API, and implement it as displayed in the following code:
    
    Tidy up
    
    To prevent undesirable charges, complete the steps in this section to clean up your resources.
    
    Delete the Amazon Bedrock Marketplace deployment
    
    If you deployed the design utilizing Amazon Bedrock Marketplace, total the following actions:
    
    1. On the Amazon Bedrock console, under Foundation designs in the navigation pane, select Marketplace deployments.
18. In the Managed releases area, find the endpoint you want to delete.
19. Select the endpoint, and on the Actions menu, select Delete.
20. Verify the endpoint details to make certain you’re deleting the proper release: 1. Endpoint name.
21. Model name.
22. Endpoint status
    
    Delete the SageMaker JumpStart predictor
    
    The SageMaker JumpStart design you deployed will sustain expenses if you leave it running. Use the following code to erase the endpoint if you wish to stop sustaining charges. For more details, see Delete Endpoints and Resources.
    
    Conclusion
    
    In this post, we checked out how you can access and release the DeepSeek-R1 model using Bedrock Marketplace and SageMaker JumpStart. Visit SageMaker JumpStart in SageMaker Studio or Amazon Bedrock Marketplace now to get begun. For more details, describe Use Amazon Bedrock tooling with Amazon SageMaker JumpStart designs, SageMaker JumpStart pretrained designs, Amazon SageMaker JumpStart Foundation Models, Amazon Bedrock Marketplace, and Starting with Amazon SageMaker JumpStart.
    
    About the Authors
    
    Vivek Gangasani is a Lead Specialist Solutions Architect for Inference at AWS. He assists emerging generative AI companies develop ingenious options utilizing AWS services and sped up compute. Currently, he is focused on developing techniques for fine-tuning and enhancing the reasoning performance of large language designs. In his spare time, Vivek takes pleasure in hiking, seeing movies, and trying different cuisines.
    
    Niithiyn Vijeaswaran is a Generative AI Specialist Solutions Architect with the Third-Party Model Science team at AWS. His area of focus is AWS AI accelerators (AWS Neuron). He holds a Bachelor’s degree in Computer Science and Bioinformatics.
    
    Jonathan Evans is an Expert Solutions Architect working on generative AI with the Third-Party Model Science team at AWS.
    
    Banu Nagasundaram leads item, engineering, and tactical partnerships for Amazon SageMaker JumpStart, SageMaker’s artificial intelligence and generative AI hub. She is enthusiastic about developing solutions that assist clients accelerate their AI journey and unlock organization value.