Cloud

Amazon Bedrock Expands OpenAI GPT-6 Access with Global Cross-Region Inference

Amazon Bedrock now provides access to OpenAI GPT-6 models with global cross-Region inference capabilities in Thailand, Malaysia, Singapore, Indonesia, and Taiwan, enabling broader deployment and usage across multiple geographic locations.

AI-generated editorial illustration of the article topic; not a photograph, product screenshot or measured result.

Global Cross-Region Inference for OpenAI GPT-6 Models

Amazon Bedrock now provides access to OpenAI GPT-6 models with global cross-Region inference capabilities in Thailand, Malaysia, Singapore, Indonesia, and Taiwan. This feature allows teams in these regions to access the latest OpenAI models beyond local capacity limits, which can be a constraint with single-Region inference. Specifically, Amazon Bedrock offers the GPT-6 Sol and GPT-6 Luna models through global cross-Region inference from all five AWS Regions, while the GPT-6 Astra model is available from the Asia Pacific (Singapore) Region. This expansion supports organizations requiring high availability and low-latency access to advanced AI models across Southeast Asia and Taiwan.

Model Availability and Access Paths

Model Availability and Access Paths: OpenAI GPT-6 models, including GPT-6 Sol, GPT-6 Luna, and GPT-6 Astra, are accessible via global cross-Region inference on Amazon Bedrock, as noted in the preceding Regions through the bedrock-runtime endpoint. This feature allows users to invoke these models from Asia Pacific (Singapore) using the Converse API, OpenAI-compatible APIs, and coding clients. The global cross-Region inference profiles for the GPT-6 family ensure that each of the three models is available through this capability, enabling broader deployment and usage across multiple geographic locations.

Inference Profiles and Region Support

Amazon Bedrock now supports global cross-Region inference for OpenAI GPT models in several Asia Pacific Regions, including Thailand, Malaysia, Singapore, Indonesia, and Taiwan. This feature routes requests across supported commercial AWS Regions where the model is deployed, providing broad access to Amazon Bedrock capacity. Available inference profiles include global.openai.gpt-6-sol, global.openai.gpt-6-luna, and global.openai.gpt-6-astra, each with specific source and destination Regions. For example, the global.openai.gpt-6-sol profile supports source Regions such as Asia Pacific (Singapore) and destination Regions like Asia Pacific (Thailand). Inference profile availability and Region support may change as AWS adds new Regions and profiles, so users are advised to confirm details on the model card before deployment. The Amazon Bedrock console lists these profiles under system-defined inference profiles.

Implementation and API Usage

Implementation of OpenAI GPT models on Amazon Bedrock in Thailand, Malaysia, Singapore, Indonesia, and Taiwan involves using the Amazon Bedrock Runtime API. Developers can invoke GPT-6 models through three access paths: the Amazon Bedrock Converse API, the OpenAI Responses API, and the OpenAI Chat Completions API. A sample implementation sets the AWS region to ap-southeast-1 and specifies a profile ID such as global.openai.gpt-6-sol. The code snippet includes logging and client configuration for the Bedrock Runtime service, with a note that AccessDeniedException may indicate an issue with the IAM policy. The model provider configuration specifies the region and uses SigV4 signing, which does not require a key rotation.

IAM Permissions and Security Considerations

When accessing OpenAI GPT models on Amazon Bedrock with global cross-Region inference in Thailand, Malaysia, Singapore, Indonesia, and Taiwan, proper IAM permissions are essential for secure and effective model invocation. According to AWS Machine Learning, an IAM policy must include four statements to enable global cross-Region inference. These include access to the inference profile and the project/default resource in the source Region. Additionally, AWS Machine Learning specifies that the Region-agnostic ARN for the foundation model, arn:aws:bedrock:::foundation-model/openai.gpt-6-sol, must be used with aws:RequestedRegion set to unspecified to allow cross-Region routing. As noted in AWS Machine Learning, this IAM policy specifically grants permission to invoke GPT-6 Sol through global cross-Region inference from ap-southeast-1. These configurations ensure secure and efficient model access across multiple Regions without requiring explicit regional permissions.