Skip to main content

​
Overview

Cerebras Inference is available through AWS Marketplace, allowing you to access our ultra-fast inference models with billing consolidated through your AWS account. This integration is ideal for organizations that want to:
  • Consolidate AI infrastructure costs with existing AWS spending
  • Leverage AWS Enterprise Discount Programs (EDP) and committed spend
  • Simplify procurement and billing processes
  • Access Cerebras models without separate vendor management
All usage is billed monthly through your AWS account as a marketplace line item, making it easy to track and manage costs alongside your other AWS services.

​
Prerequisites

Before you begin, ensure you have:
  • AWS Account - An active AWS account with permissions to subscribe to Marketplace offerings. Visit AWS Marketplace if you need to create an account.
  • Cerebras Cloud Account - You’ll create or link this during the subscription process
The email addresses for your AWS and Cerebras accounts do not need to match. You can link any existing Cerebras account to your AWS Marketplace subscription.

​
Subscribe to Cerebras on AWS Marketplace

1

Find Cerebras in AWS Marketplace

Navigate to the AWS Marketplace and locate the Cerebras Inference listing:
  1. Log in to the AWS Marketplace
  2. Search for “Cerebras” in the marketplace search bar
  3. Select the Cerebras Fast Inference Cloud listing from the search results
2

Review and subscribe

Review the pricing details and complete your subscription:
  1. On the Cerebras listing page, select View purchase options
  2. Review the pricing details and terms carefully
  3. Click Subscribe to proceed with your subscription
Make sure you understand the pricing model before subscribing. Charges are based on API usage (per million tokens) and billed monthly through your AWS account.
3

Set up your Cerebras account

After subscribing, you’ll be redirected to link or create your Cerebras account:
  1. Click Set up your account on the confirmation page
  2. You’ll be redirected to https://cloud.cerebras.ai/dashboard
  3. Either log in to your existing Cerebras account or create a new one
  4. Your Cerebras account will automatically be linked to your AWS Marketplace subscription
If you already have a Cerebras account, simply log in with your existing credentials. The system will link it to your AWS subscription automatically.
4

Generate your API key

Once logged in to your Cerebras dashboard, create an API key for authentication:
  1. Navigate to the API Keys page
  2. Click Create API Key
  3. Give your key a descriptive name (e.g., “Production API Key” or “AWS Marketplace Key”)
  4. Copy and securely store your API key - you won’t be able to see it again
Treat your API key like a password. Never commit it to version control or share it publicly.

​
Using Cerebras Inference

Once you’ve completed the subscription and setup, you can start using Cerebras Inference with your API key. For detailed integration instructions, code examples, and API documentation, visit:
When making API requests, include the header X-Cerebras-3rd-Party-Integration: aws-marketplace to ensure your usage is properly tracked for AWS Marketplace billing.

​
Available Models

Cerebras offers several state-of-the-art models through AWS Marketplace:
ModelDescriptionBest For
llama-3.3-70bMeta’s latest Llama model with 70B parametersBest for complex reasoning, long-form content, and tasks requiring deep understanding
qwen-3-32bAlibaba’s Qwen model with strong multilingual supportBalanced performance for general-purpose applications
llama3.1-8bSmaller, faster Llama modelFastest option for simple tasks and high-throughput scenarios
gpt-oss-120bLarge open-source modelLargest model for the most demanding tasks
zai-glm-4.6Advanced 357B parameter modelAdvanced 357B parameter model with strong reasoning capabilities
For the latest model availability, specifications, and performance benchmarks, visit our models documentation.

​
Account and Billing Management

​
Viewing Usage

You can monitor your API usage in two places: Cerebras Dashboard Visit the Usage page to view detailed metrics:
  • Total requests and tokens processed
  • Usage breakdown by model
  • Usage breakdown by API key
  • Real-time usage metrics and trends
  • Historical usage data
AWS Marketplace Console Access subscription details through AWS:
  1. Go to your AWS Marketplace subscriptions
  2. Select your username in the upper-right corner
  3. Choose Your AWS Marketplace Software
  4. Select Manage Subscriptions
  5. Find Cerebras Inference in your active subscriptions

​
Understanding Billing

All billing is handled through your AWS account with transparent, usage-based pricing:
  • Billing Location: AWS Billing and Cost Management console
  • Billing Frequency: Monthly, based on API usage
  • Pricing Model: Per-million tokens for inputs and outputs (varies by model)
  • AWS Integration: Charges appear as AWS Marketplace line items
  • Payment Method: Uses your existing AWS payment method
To view your Cerebras charges:
  1. In the AWS Console, select your username in the upper-right corner
  2. Choose Billing and Cost Management
  3. Navigate to Bills or Cost Explorer
  4. Look for Cerebras charges under AWS Marketplace
AWS Marketplace uses a $0.01 SKU for one-to-one conversion of API charges. For detailed pricing by model, see our pricing page.

​
Managing Your Subscription

To modify or cancel your subscription:
  1. Navigate to the AWS Marketplace console
  2. Select Manage Subscriptions
  3. Find your Cerebras Inference subscription
  4. Choose the action you want to take (modify, cancel, etc.)
Canceling your AWS Marketplace subscription will immediately revoke access to Cerebras Inference. Any in-flight API requests may fail, and you will no longer be able to make new requests.

​
Next Steps

Now that you’re set up with Cerebras through AWS Marketplace, explore these resources to get the most out of your integration:

API Reference

Complete API documentation with all endpoints and parameters

Model Documentation

Detailed model specifications, capabilities, and benchmarks

Best Practices

Optimize your inference requests for performance and cost

GLM4.6 Migration Guide

Guide for migrating to the GLM-4.6 model

​
FAQ

No, the email addresses for your AWS and Cerebras accounts do not need to match. During the subscription setup process, you can link any existing Cerebras account to your AWS Marketplace subscription, regardless of the email address used for either account.
Charges are billed monthly based on your usage. After you use the API, allow 24-48 hours for charges to appear in your AWS Billing and Cost Management console. Usage metrics are available in real-time on the Cerebras dashboard, but AWS billing is processed on a monthly cycle.
Yes! Cerebras Inference charges through AWS Marketplace typically count toward your AWS committed spend and can be covered by EDP credits. Consult with your AWS account manager to confirm how marketplace purchases apply to your specific agreement.
Canceling your subscription will immediately revoke access to Cerebras Inference. Your API keys will stop working, and any in-flight requests may fail. You will be billed for any usage up to the cancellation point. If you need to resume service, you’ll need to subscribe again through AWS Marketplace.
Yes, you can create multiple API keys within your Cerebras account. All usage across these keys will be consolidated and billed to your AWS account. This is useful for separating development, staging, and production environments or tracking usage by team or project.
Rate limits are set per account. If you need higher rate limits for production workloads, contact [email protected] to discuss your requirements. Enterprise customers can access significantly higher throughput limits.

​
Troubleshooting

Problem: Getting 401 Unauthorized errorsSolution:
  • Verify your API key is correct in your .env file
  • Ensure your AWS Marketplace subscription is active
  • Check that your Cerebras account is properly linked to AWS
  • Confirm you’re using the correct base URL: https://api.cerebras.ai/v1
  • Try generating a new API key from the Cerebras dashboard
  • Verify the X-Cerebras-3rd-Party-Integration header is set to aws-marketplace
Problem: Not seeing charges in AWS billingSolution:
  • Charges appear monthly, not in real-time
  • Check the AWS Marketplace section of your billing console
  • Allow 24-48 hours for charges to appear after usage
  • Verify your subscription is active in AWS Marketplace
  • Contact AWS Support for billing-specific questions
Problem: Unable to link Cerebras account to AWS subscriptionSolution:
  • Clear your browser cache and cookies
  • Try the setup process in an incognito/private browser window
  • Ensure you’re logged into the correct AWS account
  • Verify you completed the subscription process in AWS Marketplace
  • Try logging out of both AWS and Cerebras, then start fresh
Problem: Receiving 429 Too Many Requests errorsSolution:
  • Review our rate limits documentation
  • Implement exponential backoff in your retry logic
  • Consider distributing requests across multiple API keys
  • Monitor your usage patterns in the Cerebras dashboard
  • Contact sales to discuss higher rate limits for production workloads
Problem: Getting 404 errors when specifying a modelSolution:
  • Ensure you’re using the correct model name format: model-name
  • Verify the model is available in your region
  • Check the models documentation for current model availability
  • Some models may have limited availability - contact support for details

​
Support

For technical support and questions: Cerebras Technical Issues AWS Marketplace Issues
Ready to experience the fastest inference in the world? Start building with Cerebras through AWS Marketplace today and see the difference that purpose-built AI hardware makes.