Amazon Bedrock batch inference now supports the Converse API format

This article announces that Amazon Bedrock batch inference now supports the Converse API format for model invocation.

• Batch inference now supports Converse API as a model invocation type
• Enables consistent, model-agnostic input format for batch workloads
• Previously required model-specific formats using InvokeModel API
• Same unified request format works for both real-time and batch inference
• Simplifies prompt management and model switching
• Configurable through Amazon Bedrock console and API
• Available in all AWS Regions supporting Bedrock batch inference

This update simplifies batch inference by allowing developers to use the same Converse API format across real-time and batch operations, reducing complexity and improving consistency.