Supported Regions, models, and quotas - Amazon Bedrock
Supported RegionsSupported target modelsSupported judge modelsQuotasSupported multimodal file types
Services or capabilities described in AWS documentation might vary by Region. To see the differences applicable to the AWS European Sovereign Cloud Region, see the AWS European Sovereign Cloud User Guide.

Supported Regions, models, and quotas

Supported Regions

Advanced Prompt Optimization is available in the following AWS Regions:

  • US East (N. Virginia) (eusc-de-east-1)

  • US East (Ohio) (us-east-2)

  • US West (Oregon) (us-west-2)

  • Canada (Central) (ca-central-1)

  • South America (São Paulo) (sa-east-1)

  • Europe (Ireland) (eu-west-1)

  • Europe (London) (eu-west-2)

  • Europe (Frankfurt) (eu-central-1)

  • Europe (Zurich) (eu-central-2)

  • Asia Pacific (Mumbai) (ap-south-1)

  • Asia Pacific (Tokyo) (ap-northeast-1)

  • Asia Pacific (Seoul) (ap-northeast-2)

  • Asia Pacific (Singapore) (ap-southeast-1)

  • Asia Pacific (Sydney) (ap-southeast-2)

Supported target models

Advanced Prompt Optimization supports Amazon Bedrock models that output text modality as optimization targets. New text-output models are added as optimization targets as they become available. There can be a short delay between a model's launch on Amazon Bedrock and its availability in Advanced Prompt Optimization. The model must also be available in the Region where you create the job.

Judge models for custom LLM-as-a-judge evaluation are a separate, explicitly supported list. See the Supported judge models section on this page.

If a model you expect is not available, see Troubleshooting or contact AWS Support.

Supported judge models

When using the custom LLM-as-a-judge evaluation method, the following models are available as judge models:

  • anthropic.claude-opus-4-6-v1

  • anthropic.claude-opus-4-7

  • anthropic.claude-opus-4-8

  • anthropic.claude-opus-5

  • anthropic.claude-opus-5-5

  • anthropic.claude-sonnet-4-5-20250929-v1:0

  • anthropic.claude-sonnet-4-6

  • anthropic.claude-sonnet-5

  • anthropic.claude-sonnet-5-5

In customLLMJModelId, specify the judge model as a base model ID (for example, anthropic.claude-sonnet-4-6), as a cross-region inference profile ID for one of these models (for example, us.anthropic.claude-sonnet-4-6), or as an application inference profile ARN that routes to one of these models. A judge model that is not on this list is rejected with a ValidationException when you create the job.

Quotas

Quota Limit
Concurrent jobs 20 per account per region
Input file size 50 MB
Templates per job 10
Evaluation samples per template 100
Text variables per template 20
Multimodal files per sample 100
Models per job 5
Steering criteria per template 5

Supported multimodal file types

Category Supported formats Notes
Image JPG (.jpg), JPEG (.jpeg), PNG (.png), GIF (.gif), WebP (.webp) Support depends on the target model. For animated GIF and WebP files, models typically use only the first frame. To include multiple frames, split them into separate image files and include them as individual multimodal files in the evaluation sample.
Document PDF (.pdf)