Claude Fable 5.2 (AWS Bedrock)

Anthropic launched Claude Fable 5.2 and Claude Mythos 5.2 on November 9, 2026, saying the models answer up to 30% faster than the 5.1 generation and cost $8 per million input tokens and $40 per million output tokens, down from $10 and $50. An optional accelerated mode speeds generation up to 2.5x and holds its $16 and $80 pricing, while Fable 5.2 is available on all paid plans and major cloud platforms. Reported agentic results include one engineer asking for an operating system from an empty folder, with the model coordinating more than 1,000 subagents for eight days to produce a bootable system with its own kernel and drivers, plus a critical Python service migrated to Rust that booted 26 times faster. Mythos 5.2, limited to organizations passing Anthropic's Cyber or Life Sciences Verification Programs, found 10.2% more vulnerabilities than Opus 5.5 in an internal evaluation of 1,829 public vulnerabilities.

Amazon Bedrock and Claude Platform on AWS now serve Claude Fable 5.2 and Claude Mythos 5.2, with Amazon noting Fable 5.2 as Anthropic's most capable generally available model for technical work. The Bedrock deployment covers agentic coding across repositories, scientific research including formalizing large mathematical proofs, financial analysis, and data analysis, running for hours to days unattended while verifying its own output. Anthropic's alignment testing showed Mythos 5.2, the same underlying model as Fable 5.2 without safeguards, scored lower than Opus 5.5, Sonnet 5.5, and Mythos 5.1 on behaviors such as hallucinating inputs and falsely claiming task completion, and Anthropic expects Fable 5.2 to behave similarly. Safeguards limit misuse in cybersecurity, life sciences, and model distillation, so blocked cybersecurity or biology queries return responses from Opus 4.8 and Opus 5 respectively.

Anthropic's homepage lists Claude Fable 5.2 and Mythos 5.2 as its current flagship models, describing them as its most capable models for daunting technical work, with Fable 5.2 setting a new bar in coding, finance, and the sciences. The company publishes model details and a system card for each, alongside announcements for Opus 5.5, which leads in agentic coding and knowledge work at 40% lower cost than Opus 5. The homepage also points to Anthropic's Sonnet 5.5 release, described as a clear upgrade over Claude Sonnet 5 that runs more than 30% faster and costs up to 30% less for most workloads, alongside the Anthropic Institute, Claude's Constitution, and the company's responsible scaling policy framework.

Anthropic announced Claude Fable 5.2 on November 9, 2026, describing it as its most capable model for technical work, setting a new bar in coding, finance, and the sciences. The model is available to Pro, Max, Ultra, Team, and Enterprise users, and for developers through the Claude Platform natively, available marketplaces, and Amazon Web Services, Google Cloud, and Microsoft Foundry. Pricing drops to $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for Fable 5.1, with cache reads at $0.20 per million tokens. Developers can call it via the Claude API using the claude-fable-5-2 model id, and a fast mode delivers up to 2.5x faster speed at $16 per million input and $80 per million output tokens.