Generative AI platforms & models · AWS vs Google Cloud
Amazon SageMaker JumpStart / Amazon Bedrock Marketplace vs Model Garden
SageMaker JumpStart and Bedrock Marketplace map to Model Garden, which lists Google, partner, and open models and deploys open models to managed endpoints.
Side-by-side comparison
| Aspect | Amazon Web Services | Google Cloud |
|---|---|---|
| Service | Amazon SageMaker JumpStart Amazon Bedrock Marketplace | Model Garden |
| Lifecycle | Generally available Generally available | Generally available |
| What it is | Hub of pretrained and open-weight models plus solution templates, deployed and fine-tuned on SageMaker infrastructure. Supports private and curated hubs. Discover and deploy additional third-party and specialized models onto managed endpoints, used through Bedrock APIs. | Catalog of Google, partner, and open models. Partner models (for example Anthropic Claude, xAI Grok, Mistral) are offered as managed APIs. Open models (for example Llama, DeepSeek, Qwen, Gemma) deploy to your own endpoints. |
| Pricing model | SageMaker instance-hours for the endpoint (plus any model-provider software fee for Marketplace models). | Managed partner models are billed per token. Self-deployed open models are billed per node-hour for the endpoint's machine and accelerators. Some partner models carry their own license terms. |
| Docs | Amazon SageMaker JumpStart docs Amazon Bedrock Marketplace docs | Model Garden docs |
| Pricing page | Amazon SageMaker JumpStart pricing Amazon Bedrock Marketplace pricing | Model Garden pricing |
Key differences
- Model Garden mixes managed partner APIs (pay per token) and self-deployed open models (pay for endpoint nodes) in one catalog. AWS splits these between Bedrock Marketplace and JumpStart.
- Deployed open models run on Agent Platform inference endpoints with GPUs or TPUs; JumpStart deploys to SageMaker endpoints.
Pricing summary
Pricing models only. No prices are quoted because they change by region and over time.
- AWS
- SageMaker instance-hours for the endpoint (plus any model-provider software fee for Marketplace models). Amazon SageMaker JumpStart pricing → Amazon Bedrock Marketplace pricing →
- GCP
- Managed partner models are billed per token. Self-deployed open models are billed per node-hour for the endpoint's machine and accelerators. Some partner models carry their own license terms. Model Garden pricing →
And on Azure?
Direct match Azure's equivalent is Microsoft Foundry Models (incl. Azure OpenAI), Azure Machine Learning. AWS vs Azure comparison →
FAQ
What is the Google Cloud equivalent of Amazon SageMaker JumpStart and Amazon Bedrock Marketplace?
SageMaker JumpStart and Bedrock Marketplace map to Model Garden, which lists Google, partner, and open models and deploys open models to managed endpoints.
How does pricing compare between Amazon SageMaker JumpStart and Amazon Bedrock Marketplace and Model Garden?
AWS: SageMaker instance-hours for the endpoint (plus any model-provider software fee for Marketplace models). Google Cloud: Managed partner models are billed per token. Self-deployed open models are billed per node-hour for the endpoint's machine and accelerators. Some partner models carry their own license terms. Check the official pricing pages for current rates.
What is the Azure equivalent?
Microsoft Foundry Models (incl. Azure OpenAI) and Azure Machine Learning (direct match). SageMaker JumpStart and Bedrock Marketplace map to the Microsoft Foundry model catalog's managed-compute deployments (and the Azure Machine Learning model catalog).