AWS makes Claude Fable 5.2 available for coding and science
AWS has made Claude Fable 5.2 and Claude Mythos 5.2 available through Amazon Bedrock and Claude Platform on AWS. Fable 5.2 is positioned as Anthropic’s most capable generally available model for technical work, while Mythos 5.2 targets authorized cyber security and life sciences research.
- Claude Fable 5.2 is available on Amazon Bedrock and Claude Platform on AWS.
- Fable 5.2 targets coding, mathematics, science and other technical workloads.
- Claude Mythos 5.2 requires verification for cyber security and life sciences use.
- Enterprise Frontier Safeguards supports zero data retention for eligible customers through June 30, 2027.
- AWS provides console, API and SDK access, plus a Claude Code migration skill.
Models target long-running technical work
According to AWS, Claude Fable 5.2 is designed for difficult problems in coding, mathematics and science. The model can handle ambiguous, multi-step tasks, verify its output, identify errors and optimize its work while running unattended for hours to days.
AWS lists agentic coding across large, multi-repository codebases, long-running agents coordinating many subagents, scientific research, quantitative financial analysis, and design work for 3D, hardware and motion as potential uses for Fable 5.2.
Claude Mythos 5.2 is available through Anthropic’s verification programs for two high-risk areas: cyber security, including security engineering, red teaming and penetration testing, and life sciences research, particularly biomolecular design and research synthesis.
Safety controls and data retention requirements
AWS says Anthropic’s alignment testing found that Mythos 5.2 hallucinates less, completes tasks more reliably, surfaces important information more consistently and self-reports issues more honestly than Opus 5.5, Sonnet 5.5 and Mythos 5.1. Anthropic expects Fable 5.2 to behave similarly, according to the AWS post.
Anthropic designates Fable 5.2 as a Covered Model. By default, using it requires retention for up to 30 days and human safety review. On Amazon Bedrock, Amazon personnel perform that review within the AWS boundary, and customers must set the aws_review data retention mode on their account or project before invoking the model.
Enterprise Frontier Safeguards, developed by Anthropic and AWS, combines zero data retention with safeguards designed to detect misuse. AWS says the service is available now. A version that keeps data in a customer-controlled bucket and under customer-controlled keys has not launched yet. Eligible customers can use Fable 5.2 with zero data retention through June 30, 2027.
How developers can access Claude Fable 5.2
Users can open the Amazon Bedrock console, choose Test and Playground, and select Fable 5.2. Developers can also call it through the Anthropic Messages API on bedrock-runtime, or continue using the Invoke and Converse APIs through the AWS Command Line Interface and AWS SDKs.
AWS lists an active AWS account with Amazon Bedrock access, the AWS CLI, Python 3.9 or later, Boto3 and the required IAM permissions as prerequisites for its code examples. The Bedrock model ID shown in the example is global.anthropic.claude-fable-5-2.
Teams migrating existing Claude applications can use the open-source claude-api skill in Claude Code. AWS says it can update model references, move suitable configurations to adaptive thinking, remove outdated parameters, suggest effort levels and adjust settings for deployment platforms including Amazon Bedrock. The skill can be run with /claude-api migrate.
Both models are available today on Amazon Bedrock through US Geo CRIS and Global CRIS inference profiles, with the full supported region list in the Bedrock documentation. They are also available through Claude Platform on AWS in every region where that service is offered.
AWS: Introducing Claude Fable 5.2 on AWS
