Maintained with ☕️ by
IcePanel logo

Amazon Bedrock now supports 1-hour duration for prompt caching

Share

Services

Amazon Bedrock now supports a 1-hour time-to-live (TTL) option for prompt caching for select Anthropic Claude models. With this update, you can extend the persistence of cached prompt prefixes from the default 5 minutes to 1 hour, improving cost efficiency and performance for long-running agentic workflows and multi-turn conversations. Previously, cached content remained active for a fixed 5-minute window and refreshed when reused. With the new 1-hour TTL option, you can maintain context for users who interact less frequently, or for complex agents that require more time between steps—such as tool use, retrieval, and orchestration. The 1-hour TTL is also useful for longer sessions and batch processing where you want cached content to persist across extended periods. 1-hour TTL prompt caching is generally available for Anthropic’s Claude Sonnet 4.5, Claude Haiku 4.5, and Claude Opus 4.5 in all commercial [AWS Regions](https://docs.aws.amazon.com/general/latest/gr/bedrock.html) and AWS GovCloud (US) Regions where these models are available. The 1-hour cache is billed at a different rate than the standard 5-minute cache. To learn more, refer to the [Amazon Bedrock documentation](https://docs.aws.amazon.com/bedrock/latest/userguide/prompt-caching.html) and [Amazon Bedrock Pricing](https://aws.amazon.com/bedrock/pricing/) page.